Rice three-dimensional seedling raising temperature and humidity control method and device
By constructing a PID adjustment model and a Gaussian process Bayesian model, and combining the particle swarm algorithm to optimize hyperparameters, the precise temperature and humidity control in the rice plant is achieved, solving the problem of inefficiency in traditional adjustment methods and improving production efficiency and land utilization.
Patent Information
- Application Number
- CN202510691953.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-27
- Publication Date
- 2025-07-22
AI Technical Summary
The environmental regulation of traditional rice plant relies on manual experience and is difficult to achieve real-time, precise and even control, resulting in low production efficiency.
Build a PID adjustment model and combine the Gaussian process Bayesian model and particle swarm algorithm to optimize hyperparameters, establish an environmental control model with adaptive learning capabilities, and accurately regulate temperature and humidity through ultrasonic humidifiers, air-cooling components and heating components.
Real-time, accurate and even control of temperature and humidity in the rice seedling plant has been achieved, which has improved production efficiency and improved land utilization.
Smart Images
Figure CN120353283A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of seedling raising equipment, and relates to a method and device for controlling temperature and humidity in three-dimensional rice seedling raising. Background Art
[0002] As one of the important food crops in China, the seedling raising link is the key to ensuring stable and high yields. However, traditional seedling raising methods are greatly affected by environmental and climatic conditions, with low land utilization rate, long seedling raising cycle, and inability to guarantee the quality of seedlings, which has become a "bottleneck" problem restricting rice production. Therefore, three-dimensional rice seedling raising factories have emerged. They can cultivate more seedlings under the same seedling raising area, and rely on a circulating drive system to achieve the dynamic distribution and movement of seedlings in a three-dimensional space, ensuring that each layer of seedlings obtains suitable growth conditions and improving the utilization rate of seedling raising space resources. This not only effectively solves the problem of rice seedling production but also ensures the unified supply of seedlings.
[0003] However, due to the combined effects of multi-parameter coupling (such as temperature, humidity, light, CO2 concentration, soil humidity, etc.), uneven spatial distribution, rapid dynamic changes, and external interference factors in three-dimensional rice seedling raising factories, the environment in these factories is complex and changeable. The existing environmental regulation in three-dimensional rice seedling raising factories mainly relies on manual experience, making it difficult to regulate the environment in real time, accurately, and uniformly, thus reducing the production efficiency of three-dimensional rice seedling raising factories. Summary of the Invention
[0004] The purpose of the present invention is to provide a method and device for controlling temperature and humidity in three-dimensional rice seedling raising, which can regulate the environment in real time, accurately, and uniformly, and improve the production efficiency of three-dimensional rice seedling raising factories.
[0005] To achieve the above purpose, the technical solution provided by the present invention is as follows: A method for controlling temperature and humidity in three-dimensional rice seedling raising includes the following steps: Construct a PID adjustment model for controlling temperature and humidity in a three-dimensional rice seedling raising factory, adjust the proportional parameter, integral parameter, and differential parameter in the PID adjustment model according to a set gradient, and obtain the expected value of temperature, the final value of temperature, and the deviation value of temperature in the three-dimensional rice seedling raising factory, as well as the corresponding PID control parameters, and the expected value of humidity, the final value of humidity, and the deviation value of humidity, and the corresponding PID control parameters.
[0006] Preprocess the expected value of temperature, the final value of temperature, and the deviation value of temperature, and the corresponding PID control parameters, and the expected value of humidity, the final value of humidity, and the deviation value of humidity, and the corresponding PID control parameters to obtain training data.
[0007] Build a Gaussian process Bayesian model, introduce the particle swarm optimization algorithm to optimize the hyperparameters of the Gaussian process Bayesian model, and use the optimized hyperparameters of the Gaussian process Bayesian model to establish a Gaussian process Bayesian control model with adaptive learning ability.
[0008] Input the training data into the Gaussian process Bayesian control model for training to obtain an environmental control model that can predict the PID control parameters in real time.
[0009] Collect the temperature and humidity in the three-dimensional rice seedling raising factory in real time. Input the current value of the temperature, the target value of the temperature, the current value of the humidity, and the target value of the humidity in the three-dimensional rice seedling raising factory into the environmental control model. The environmental control model outputs the temperature PID control parameters and the humidity PID control parameters, and then input the temperature PID control parameters and the humidity PID control parameters into the PID adjustment model to realize the control of the temperature and humidity in the three-dimensional rice seedling raising factory.
[0010] The features of the present invention also lie in: When preprocessing the expected value of temperature, the final value of temperature, the deviation value of temperature and the corresponding PID control parameters, and the expected value of humidity, the final value of humidity, the deviation value of humidity and the corresponding PID control parameters, a normalization processing method is adopted. The expected value of temperature, the final value of temperature, the deviation value of temperature and the corresponding PID control parameters, and the expected value of humidity, the final value of humidity, the deviation value of humidity and the corresponding PID control parameters are processed into values within the interval [0, 1] through the normalization formula.
[0011] When using the particle swarm optimization algorithm to optimize the hyperparameters of the Gaussian process Bayesian model, first select the fitness function of the particle swarm optimization algorithm, solve the fitness function of the particle swarm optimization algorithm, and then use the particle swarm optimization algorithm to optimize the hyperparameters of the Gaussian process Bayesian model.
[0012] When establishing the Gaussian process Bayesian control model, set the input and output of the model. The input is the expected value of temperature, the final value of temperature, the deviation value of temperature, the expected value of humidity, the final value of humidity, and the deviation value of humidity, and the output is the parameters of temperature control and the parameters of humidity control.
[0013] When inputting the training data into the Gaussian process Bayesian control model for training to obtain the environmental control model, first input part of the training data into the Gaussian process Bayesian control model for training, and then input the remaining training data into the environmental control model for verification.
[0014] A three-dimensional rice seedling raising temperature and humidity control device, comprising: An ultrasonic humidifier for adjusting the humidity in the three-dimensional rice seedling raising factory; A plurality of first air-cooling components for generating a horizontal air flow to reduce the temperature in the three-dimensional rice seedling raising factory; A plurality of second air-cooling components for generating vertical airflows to reduce the temperature inside the vertical rice seedling raising factory; A plurality of heating components for increasing the temperature inside the vertical rice seedling raising factory; A controller electrically connected to the ultrasonic humidifier, the plurality of first air-cooling components, the plurality of second air-cooling components, and the plurality of heating components respectively. The environmental control model and the PID regulation model are set in the controller. The current value of the temperature, the target value of the temperature, the current value of the humidity, and the target value of the humidity inside the vertical rice seedling raising factory are input into the controller. The controller controls the start and stop of the ultrasonic humidifier, the plurality of first air-cooling components, the plurality of second air-cooling components, and the plurality of heating components to control the temperature and humidity inside the vertical rice seedling raising factory.
[0015] One side of the ultrasonic humidifier is provided with a water storage tank. The water storage tank is connected to the ultrasonic humidifier through a shunt pipe. One side of the shunt pipe is provided with a plurality of water pipes. One end of each of the plurality of water pipes is of a closed structure, and the other ends of the plurality of water pipes are respectively connected to the shunt pipe. A plurality of water holes are evenly opened on each water pipe.
[0016] A rice vertical seedling raising temperature and humidity control method and device of the present invention has the following advantages: First, the present invention first constructs a PID regulation model for temperature and humidity control inside the vertical rice seedling raising factory, adjusts the proportional parameter, integral parameter, and differential parameter in the PID regulation model according to the set gradient, obtains the expected value of the temperature, the final value of the temperature, the deviation value of the temperature, and the corresponding PID control parameters, and the expected value of the humidity, the final value of the humidity, the deviation value of the humidity, and the corresponding PID control parameters inside the vertical rice seedling raising factory. Then, preprocess the expected value of the temperature, the final value of the temperature, the deviation value of the temperature, and the corresponding PID control parameters, and the expected value of the humidity, the final value of the humidity, the deviation value of the humidity, and the corresponding PID control parameters to obtain training data. Then, construct a Gaussian process Bayesian model, introduce a particle swarm optimization algorithm to optimize the hyperparameters of the Gaussian process Bayesian model, establish a Gaussian process Bayesian control model with adaptive learning ability using the optimized hyperparameters of the Gaussian process Bayesian model, and input the training data into the Gaussian process Bayesian control model for training to obtain an environmental control model that can predict PID control parameters in real time. Finally, collect the temperature and humidity inside the vertical rice seedling raising factory in real time, input the current value of the temperature, the target value of the temperature, the current value of the humidity, and the target value of the humidity inside the vertical rice seedling raising factory into the environmental control model. The environmental control model outputs the temperature PID control parameters and the humidity PID control parameters, and then input the temperature PID control parameters and the humidity PID control parameters into the PID regulation model to achieve precise control of the temperature and humidity inside the vertical rice seedling raising factory. The present invention can perform real-time, accurate, and uniform regulation of the temperature and humidity inside the vertical rice seedling raising factory, improving the production efficiency of the vertical rice seedling raising factory.
[0017] Second, the present invention realizes the precise control of temperature and humidity during rice seedling raising by innovatively integrating the particle swarm optimization algorithm, Gaussian process Bayesian model, and PID control regulation. At the same time, by adopting an advanced regional three-dimensional seedling raising management mode, the land utilization rate is effectively improved, and the problem of high cost of sensor equipment in traditional control methods is solved, promoting the development of rice seedling raising technology towards a more efficient and precise direction. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] Figure 1 It is a schematic diagram of the overall process of the present invention.
[0019] Figure 2 It is a schematic diagram of the overall structure of the present invention.
[0020] Figure 3 It is a schematic diagram of the side view structure of the present invention.
[0021] Figure 4 It is a schematic diagram of the front view structure of the present invention.
[0022] Figure 5 It is a schematic diagram of the top view structure of the present invention.
[0023] Reference numerals: 1. Ultrasonic humidifier, 2. Water pipe, 3. First air cooling component, 4. Heating component, 5. Controller, 6. Reservoir, 7. Second air cooling component. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0024] Hereinafter, the technical solutions in the present invention will be clearly and elaborately described in conjunction with the drawings. Among them, in the description of the embodiments of the present invention, unless otherwise specified, " / " means "or". For example, A / B may mean A or B: "and / or" in the text is merely a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B may mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of the present invention, "a plurality" means two or more than two. The following terms "first" and "second" are only for descriptive purposes and cannot be construed as implying or suggesting relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of such features.
[0025] As Figure 1 shown, the present invention provides a method for controlling the temperature and humidity of three-dimensional rice seedling raising, which includes the following steps: Build a PID adjustment model for temperature and humidity control in a three-dimensional rice seedling raising factory, adjust the proportional parameter, integral parameter and differential parameter in the PID adjustment model according to the set gradient, and obtain the expected value of temperature, the final value of temperature and the deviation value of temperature in the three-dimensional rice seedling raising factory, as well as the corresponding PID control parameters, and the expected value of humidity, the final value of humidity and the deviation value of humidity, and the corresponding PID control parameters.
[0026] Preprocess the expected value of temperature, the final value of temperature and the deviation value of temperature, and the corresponding PID control parameters, and the expected value of humidity, the final value of humidity and the deviation value of humidity, and the corresponding PID control parameters to obtain training data.
[0027] Build a Gaussian process Bayesian model, introduce the particle swarm optimization algorithm to optimize the hyperparameters of the Gaussian process Bayesian model, and establish a Gaussian process Bayesian control model with adaptive learning ability using the hyperparameters of the optimized Gaussian process Bayesian model.
[0028] Input the training data into the Gaussian process Bayesian control model for training to obtain an environment control model that can predict PID control parameters in real time.
[0029] Collect the temperature and humidity in the three-dimensional rice seedling raising factory in real time, input the current value of temperature, the target value of temperature, the current value of humidity and the target value of humidity in the three-dimensional rice seedling raising factory into the environment control model. The environment control model outputs the temperature PID control parameters and the humidity PID control parameters, and then input the temperature PID control parameters and the humidity PID control parameters into the PID adjustment model to realize the control of temperature and humidity in the three-dimensional rice seedling raising factory.
[0030] Among them, PID adjustment is a feedback control method widely used in industrial control systems. PID represents Proportional, Integral and Derivative. By combining these three parts, the output of the system is adjusted to reach the desired set value.
[0031] Among them, the expected value of temperature is the set target temperature value in the three-dimensional rice seedling raising factory, the final value of temperature is the actual temperature value reached after being controlled by the PID adjustment model, the deviation value of temperature is the difference between the expected value and the final value of temperature, the expected value of humidity is the set target humidity value in the three-dimensional rice seedling raising factory, the final value of humidity is the actual humidity value reached after being controlled by the PID adjustment model, and the deviation value of humidity is the difference between the expected value and the final value of humidity.
[0032] In summary, the present invention first constructs a PID adjustment model for temperature and humidity control in a three-dimensional rice seedling raising factory, adjusts the proportional parameter, integral parameter, and differential parameter in the PID adjustment model according to a set gradient, obtains the expected value of temperature, the final value of temperature, the deviation value of temperature, and the corresponding PID control parameters in the three-dimensional rice seedling raising factory, as well as the expected value of humidity, the final value of humidity, the deviation value of humidity, and the corresponding PID control parameters. Then, preprocesses the expected value of temperature, the final value of temperature, the deviation value of temperature, and the corresponding PID control parameters, and the expected value of humidity, the final value of humidity, the deviation value of humidity, and the corresponding PID control parameters to obtain training data. Then constructs a Gaussian process Bayesian model, introduces a particle swarm optimization algorithm to optimize the hyperparameters of the Gaussian process Bayesian model, uses the hyperparameters of the optimized Gaussian process Bayesian model to establish a Gaussian process Bayesian control model with adaptive learning ability, and inputs the training data into the Gaussian process Bayesian control model for training to obtain an environmental control model capable of real-time predicting PID control parameters. Finally, real-time collects the temperature and humidity in the three-dimensional rice seedling raising factory, inputs the current value of temperature, the target value of temperature, the current value of humidity, and the target value of humidity in the three-dimensional rice seedling raising factory into the environmental control model. The environmental control model outputs the temperature PID control parameters and the humidity PID control parameters. At this time, the obtained temperature PID control parameters and humidity PID control parameters are the optimal temperature PID control parameters and the optimal humidity PID control parameters. Then inputs the temperature PID control parameters and the humidity PID control parameters into the PID adjustment model to realize the control of temperature and humidity in the three-dimensional rice seedling raising factory, so that the temperature and humidity in the three-dimensional rice seedling raising factory reach the optimal. The present invention can perform real-time, accurate, and uniform regulation of the temperature and humidity in the three-dimensional rice seedling raising factory, and improves the production efficiency of the three-dimensional rice seedling raising factory.
[0033] Among them, when adjusting the proportional parameter, integral parameter, and differential parameter in the PID adjustment model according to a set gradient, the initial control parameters of temperature are set as: the proportional parameter is 3.0, the integral parameter is 0.05, and the differential parameter is 0.4; the initial control parameters of humidity are set as: the proportional parameter is 3.0, the integral parameter is 0.05, and the differential parameter is 0.4; the set gradient is: the gradient adjustment range of the proportional parameter is 1 to 5, the gradient adjustment range of the integral parameter is 0.01 to 0.1, and the gradient adjustment range of the differential parameter is 0 to 1.
[0034] Among them, when preprocessing the expected value of temperature, the final value of temperature, the deviation value of temperature and the corresponding PID control parameters, and the expected value of humidity, the final value of humidity, the deviation value of humidity and the corresponding PID control parameters, a normalization processing method is adopted. The expected value of temperature, the final value of temperature, the deviation value of temperature and the corresponding PID control parameters, and the expected value of humidity, the final value of humidity, the deviation value of humidity and the corresponding PID control parameters are processed into values within the interval [0, 1] through the normalization formula.
[0035] Among them, when optimizing the hyperparameters of the Gaussian process Bayesian model using the particle swarm optimization algorithm, first select the fitness function of the particle swarm optimization algorithm, solve the fitness function of the particle swarm optimization algorithm, and then use the particle swarm optimization algorithm to optimize the hyperparameters of the Gaussian process Bayesian model.
[0036] Specifically, the fitness function is the MAS function, which is used to evaluate the prediction performance of the Gaussian process Bayesian model. The specific formula is as follows: .
[0037] In the formula, is the predicted value of the model at time t . is the actual value at time t , n is the number of samples within the prediction period, m is the length of the seasonal cycle (for non-seasonal data, it can be set to 1), is the reference value for normalization, usually selected as the difference between the actual value of the previous period and the actual value before the seasonal cycle.
[0038] Specifically, the specific steps for solving the fitness function of the particle swarm optimization algorithm and using the particle swarm optimization algorithm to optimize the hyperparameters of the Gaussian process Bayesian model are as follows: Initialize the parameters of the particle swarm optimization algorithm, determine the learning factor and the particle swarm size parameter, and take the dimension of the entire solution space in the particle swarm as the number of parameters to be optimized in the Gaussian process Bayesian model.
[0039] Iteratively solve the fitness function, and calculate the prediction error of the Gaussian process Bayesian model according to the fitness function.
[0040] Compare within the entire particle swarm to obtain the global extreme value, compare each particle with itself to obtain the individual extreme value, compare the global extreme value with the currently calculated optimal individual extreme value, take the extreme value with a smaller fitness as the new global extreme value, and take the one with a smaller fitness value as the new individual extreme value.
[0041] Continuously update the velocity parameters and displacement parameters of the particle swarm. When the number of iterations reaches the maximum number or the error of the fitness function meets the requirements, stop the iteration. The final optimal solution of the particle swarm optimization algorithm is used as the optimal hyperparameters of the Gaussian process Bayesian model.
[0042] Among them, when establishing the Gaussian process Bayesian control model, set the input and output of the model. The input is the expected value of temperature, the final value of temperature, the deviation value of temperature, the expected value of humidity, the final value of humidity, and the deviation value of humidity. The output is the parameters for temperature control and the parameters for humidity control.
[0043] Among them, when inputting the training data into the Gaussian process Bayesian control model for training to obtain the environmental control model, first input part of the training data into the Gaussian process Bayesian control model for training, and then input the remaining training data into the environmental control model for verification.
[0044] Such as Figure 2 、 Figure 3 、 Figure 4 、 Figure 5 As shown, the present invention also provides a temperature and humidity control device for three-dimensional rice seedling raising, including an ultrasonic humidifier 1, a plurality of first air-cooling components 3, a plurality of heating components 4, a plurality of second air-cooling components 7 and a controller 5. The ultrasonic humidifier 1 is arranged on the three-dimensional seedling raising rack. The ultrasonic humidifier 1 is used to adjust the humidity inside the three-dimensional seedling raising rack. The plurality of first air-cooling components 3 are arranged on one side of the three-dimensional seedling raising rack and are evenly arranged along the length direction of the three-dimensional seedling raising rack. The plurality of first air-cooling components 3 are used to generate horizontal airflows to cool the three-dimensional seedling raising rack. The plurality of heating components 4 are arranged on one side of the three-dimensional seedling raising rack and are evenly arranged along the length direction of the three-dimensional seedling raising rack. The plurality of heating components 4 are used to heat the three-dimensional seedling raising rack. The plurality of second air-cooling components 7 are evenly arranged on the inner top of the three-dimensional seedling raising rack. The plurality of second air-cooling components 7 are used to generate vertical airflows to cool the three-dimensional seedling raising rack. The controller 5 is arranged on the three-dimensional seedling raising rack. The controller 5 is electrically connected to the ultrasonic humidifier 1, the plurality of first air-cooling components 3, the plurality of second air-cooling components 7, and the plurality of heating components 4 respectively. The environmental control model and the PID adjustment model are arranged in the controller 5. Input the current value of temperature, the target value of temperature, the current value of humidity, and the target value of humidity in the three-dimensional rice seedling raising factory into the controller 5. The controller 5 controls the start and stop of the ultrasonic humidifier 1, the plurality of first air-cooling components 3, the plurality of second air-cooling components 7, and the plurality of heating components 4 to control the temperature and humidity in the three-dimensional rice seedling raising factory.
[0045] Among them, each first air-cooling component and each second air-cooling component 7 are fans, and each heating component 4 is a graphene heating plate.
[0046] Such as Figure 2 、 Figure 3As shown in the figure, a water reservoir 6 is provided on one side of the ultrasonic humidifier 1. The water reservoir 6 is connected to the ultrasonic humidifier 1 through a shunt pipe. The small water reservoir 6 is arranged at one end of the three-dimensional seedling raising rack. A plurality of water pipes 2 are arranged on one side of the shunt pipe. The plurality of water pipes 2 are arranged at the inner bottom of the three-dimensional seedling raising rack and are evenly arranged along its length direction. One end of the plurality of water pipes 2 is a closed structure, and the other ends of the plurality of water pipes 2 are respectively connected to the shunt pipe. A plurality of water holes are evenly opened on each water pipe 2, and accurate and uniform humidification effects are achieved through the plurality of water holes. In addition, the controller 5 can intelligently adjust the control parameters of the humidifier according to the rotation speed of the three-dimensional seedling raising tray, so as to ensure a stable humidity environment at different rotation speeds.
[0047] Among them, the controller 5 adopts a high-performance industrial control computer. With its powerful data processing ability and excellent stability, this computer can efficiently run the particle swarm optimization algorithm to optimize the Gaussian process Bayesian model parameters. In order to ensure the accurate execution of instructions and the stable exchange of data, a PLC control cabinet with Modbus (serial communication protocol) communication function is also equipped. At the same time, in order to capture the environmental data in the three-dimensional seedling raising rack in real time and accurately, a plurality of temperature and humidity sensors are arranged in the three-dimensional seedling raising rack. The plurality of temperature and humidity sensors transmit the collected data to the control computer through the Modbus protocol. After receiving these data, the computer will compare them with the preset environmental parameter values and input them into the environmental control model to obtain the optimal PID control parameters. Subsequently, these parameters will be transmitted back to the PLC control cabinet, and the PLC control cabinet will realize the refined regulation of the environment of the three-dimensional seedling raising rack.
[0048] Embodiment 1 Step 1, data collection: Install sensors in the three-dimensional seedling raising factories in the target area (including multiple three-dimensional rice seedling raising factories) to monitor and record meteorological environment data in real time, such as outdoor temperature, humidity, wind speed, wind direction, solar radiation, etc. The data collection system regularly collects these meteorological data and stores them in the central database. Determine the observed three-dimensional seedling raising rack and collect its structural parameters. Conduct detailed measurements on the three-dimensional seedling raising racks in each three-dimensional rice seedling raising factory and record their structural parameters, such as height, width, number of layers, spacing between each layer, material type, etc.
[0049] Step 2, select the observed three-dimensional seedling raising rack: Based on the structural parameters of the three-dimensional seedling raising rack and the layout and location of each three-dimensional seedling raising factory, select representative three-dimensional seedling raising racks as the observation objects. These observed three-dimensional seedling raising racks should be able to reflect the temperature and humidity change trends of the entire area.
[0050] Step 3: Construct a PID regulation model related to the temperature and humidity in the three-dimensional rice seedling raising factory. Adjust the control parameters in the PID regulation model, including proportional, integral, and differential, according to the set gradient. Record and obtain the expected value of the temperature, the final value of the temperature, and the deviation value of the temperature in the three-dimensional rice seedling raising factory, as well as the corresponding PID control parameters, and the expected value of the humidity, the final value of the humidity, and the deviation value of the humidity, and the corresponding PID control parameters. The test time is one hour, and data is recorded every minute. Set the temperature during the first complete leaf elongation stage to 22°C - 26°C, the weaning stage to 22°C - 24°C, the transplanting area to 18°C - 20°C, and the relative humidity to 70% - 80%. Then record the set value, stable value, and deviation value of the temperature and humidity.
[0051] Step 4: Preprocess the expected value of the temperature, the final value of the temperature, and the deviation value of the temperature, as well as the corresponding PID control parameters, and the expected value of the humidity, the final value of the humidity, and the deviation value of the humidity, and the corresponding PID control parameters to obtain training data. Process the data into values within the interval [0, 1] through the normalization formula. Determine and solve the fitness function of the particle swarm algorithm, specifically the MAS function, and use the particle swarm algorithm to optimize the hyperparameters of the Gaussian process Bayesian model.
[0052] Step 5: Construct a Gaussian process Bayesian model. Introduce the particle swarm algorithm to optimize the hyperparameters of the Gaussian process Bayesian model. Use the optimized hyperparameters of the Gaussian process Bayesian model to establish a Gaussian process Bayesian control model. Specifically, solve the fitness function of the particle swarm algorithm. The dimension of the entire solution space in the particle swarm is the total number of hyperparameters in the Bayesian. Determine the learning factor and the particle swarm size parameter, perform iterative solution on the fitness function, subtract the output result of the network from the expected value and take the absolute value, then find the individual extreme value and the population extreme value, and continuously update the particle swarm velocity parameter and displacement parameter. Stop the iteration when the number of iterations reaches the maximum number or the error of the fitness function meets the requirements. The final optimal solution of the particle swarm algorithm is used as the optimal hyperparameter combination of the Gaussian process Bayesian model.
[0053] Step 6: Use the optimal hyperparameter combination obtained by the particle swarm algorithm to establish a Gaussian process Bayesian control model, and use the training data to train the Gaussian process Bayesian control model to obtain an environment control model. Real-time collect the temperature and humidity in the three-dimensional rice seedling raising factory. Input the current value of the temperature, the target value of the temperature, the current value of the humidity, and the target value of the humidity in the three-dimensional rice seedling raising factory into the controller 5. The environment control model inside the controller 5 outputs the temperature PID control parameters and the humidity PID control parameters, and sends them to the PID regulation model. The PID regulation model drives the controller 5 to control the ultrasonic humidifier 1, multiple first air-cooling components 3, heating component 4, small reservoir 6, and second air-cooling component 7 to achieve the temperature and humidity in the three-dimensional rice seedling raising factory.
[0054] It will be understood that the present invention is described by way of some embodiments, and those skilled in the art will appreciate that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the present invention. Additionally, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific circumstances and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of the present invention are within the scope protected by the present invention.
Claims
1. A method for controlling temperature and humidity in three-dimensional rice seedling raising, characterized in that, The method includes the following steps: Construct a PID regulation model for temperature and humidity control in a three-dimensional rice seedling raising factory, adjust the proportional parameter, integral parameter and differential parameter in the PID regulation model according to the set gradient, and obtain the expected value of temperature, the final value of temperature and the deviation value of temperature in the three-dimensional rice seedling raising factory, as well as the corresponding PID control parameters, and the expected value of humidity, the final value of humidity and the deviation value of humidity, and the corresponding PID control parameters; Preprocess the expected value of temperature, the final value of temperature and the deviation value of temperature, and the corresponding PID control parameters, and the expected value of humidity, the final value of humidity and the deviation value of humidity, and the corresponding PID control parameters to obtain training data; Construct a Gaussian process Bayesian model, introduce a particle swarm optimization algorithm to optimize the hyperparameters of the Gaussian process Bayesian model, and establish a Gaussian process Bayesian control model with adaptive learning ability using the optimized hyperparameters of the Gaussian process Bayesian model; Input the training data into the Gaussian process Bayesian control model for training to obtain an environmental control model that can predict PID control parameters in real time; Collect the temperature and humidity in the three-dimensional rice seedling raising factory in real time, input the current value of temperature, the target value of temperature, the current value of humidity and the target value of humidity in the three-dimensional rice seedling raising factory into the environmental control model, the environmental control model outputs the temperature PID control parameters and the humidity PID control parameters, and then input the temperature PID control parameters and the humidity PID control parameters into the PID regulation model to realize the control of temperature and humidity in the three-dimensional rice seedling raising factory.
2. The method for controlling temperature and humidity in three-dimensional rice seedling raising according to claim 1, characterized in that When preprocessing the expected value of temperature, the final value of temperature and the deviation value of temperature, and the corresponding PID control parameters, and the expected value of humidity, the final value of humidity and the deviation value of humidity, and the corresponding PID control parameters, adopt a normalization processing method, and process the expected value of temperature, the final value of temperature and the deviation value of temperature, and the corresponding PID control parameters, and the expected value of humidity, the final value of humidity and the deviation value of humidity, and the corresponding PID control parameters into values within the interval [0, 1] through the normalization formula.
3. The rice three-dimensional seedling raising temperature and humidity control method according to claim 1, wherein, When using the particle swarm optimization algorithm to optimize the hyperparameters of the Gaussian process Bayesian model, first select the fitness function of the particle swarm optimization algorithm, solve the fitness function of the particle swarm optimization algorithm, and then use the particle swarm optimization algorithm to optimize the hyperparameters of the Gaussian process Bayesian model.
4. The method for controlling temperature and humidity in three-dimensional rice seedling raising according to claim 1, wherein, When establishing a Gaussian process Bayesian control model, set the input and output of the model. The input is the expected value of temperature, the final value of temperature, the deviation value of temperature, the expected value of humidity, the final value of humidity and the deviation value of humidity, and the output is the parameters for temperature control and the parameters for humidity control.
5. The method for controlling the temperature and humidity of three-dimensional rice seedling raising according to claim 1, characterized in that, When inputting the training data into the Gaussian process Bayesian control model for training to obtain an environmental control model, first input part of the training data into the Gaussian process Bayesian control model for training, and then input the remaining training data into the environmental control model for verification.
6. A temperature and humidity control device for three-dimensional rice seedling raising, characterized in that, The control method according to claim 1 includes: An ultrasonic humidifier (1) for adjusting the humidity in the three-dimensional rice seedling raising factory; A plurality of first air-cooling components (3) for generating a horizontal air flow to reduce the temperature in the three-dimensional rice seedling raising factory; A plurality of second air-cooling components (7) for generating vertical airflows to reduce the temperature inside the rice vertical seedling raising factory; A plurality of heating components (4) for increasing the temperature inside the rice vertical seedling raising factory; A controller (5) electrically connected to the ultrasonic humidifier (1), a plurality of first air-cooling components (3), a plurality of second air-cooling components (7), and a plurality of heating components (4) respectively. The environmental control model and the PID adjustment model are set inside the controller (5). The current value of the temperature, the target value of the temperature, the current value of the humidity, and the target value of the humidity inside the rice vertical seedling raising factory are input into the controller (5). The controller (5) controls the start and stop of the ultrasonic humidifier (1), a plurality of first air-cooling components (3), a plurality of second air-cooling components (7), and a plurality of heating components (4) to control the temperature and humidity inside the rice vertical seedling raising factory.
7. The rice three-dimensional seedling raising temperature and humidity control device according to claim 6, characterized in that, A water storage tank (6) is arranged on one side of the ultrasonic humidifier (1). The water storage tank (6) is connected to the ultrasonic humidifier (1) through a shunt pipe. A plurality of water pipes (2) are arranged on one side of the shunt pipe. One end of each of the plurality of water pipes (2) is of a closed structure, and the other ends of the plurality of water pipes (2) are respectively connected to the shunt pipe. A plurality of water holes are evenly formed in each of the water pipes (2).