Intelligent photogroup illuminator

Abstract

The invention belongs to the technical field of optical radiation measurement, and discloses an intelligent photogroup illuminometer which comprises a multi-sensor module, an MCU microcontroller module, a BP algorithm center module, an LED display module and a handheld shell. Based on the above modules, the specific use method of the illuminometer comprises the following steps: S1, equipment preparation and calibration; S2, measurement parameter setting; S3, data acquisition; S4, data preprocessing; the intelligent photogroup illuminator is combined with a high-precision sensor and an advanced data analysis technology, remarkable benefits are brought to agricultural production, and by accurately measuring key parameters such as illuminance, temperature and gas concentration of chlorophyll absorption wavelength, the intelligent photogroup illuminator can reflect the growth environment condition of crops in real time, so that the agricultural production quality is improved. And the data preprocessing function ensures the analysis quality.

Description

Technical Field

[0001] The invention belongs to the technical field of light radiation measurement, and in particular relates to an intelligent photosynthetic illuminance meter. Background Art

[0002] The intelligent photosynthetic illuminance meter is an advanced instrument specially designed to measure parameters related to photosynthesis. It combines modern sensing technology, microelectronics technology and data processing technology, and can monitor key parameters such as light intensity, temperature and carbon dioxide concentration in the environment of plant leaves in real time. These parameters are of great significance for evaluating photosynthesis efficiency, understanding plant growth conditions and optimizing agricultural production environment. The existing intelligent photosynthetic illuminance meter integrates high-precision sensors and complex data analysis technology. Although it improves functionality and accuracy, it also brings problems of technical complexity and high cost, which constitutes an economic burden for small farmers or economically underdeveloped areas. Sensors are easily affected by environmental factors such as temperature, humidity and dust, which may lead to inaccurate data and affect the assessment of the crop growth environment. In addition, the construction and training of the BP neural network model requires deep expertise and a large amount of historical data, which is a major challenge for non-professional users. Although the remote data transmission function provides convenience, it is also accompanied by data security risks, and effective security measures need to be taken. The technical threshold and operational complexity of intelligent photosynthetic illuminance meters limit their user groups. Therefore, it is crucial to provide a user-friendly interface, detailed operating guides and effective training support to improve user acceptance and promote the widespread application of technology. Summary of the invention

[0003] The purpose of the present invention is to provide an intelligent photosynthetic illuminance meter to solve the problems raised in the above background technology.

[0004] In order to achieve the above object, the present invention provides the following technical solutions: an intelligent photosynthetic illuminance meter, the illuminance meter package comprises a multi-sensor module, an MCU microcontroller module, a BP algorithm center module, an LED display module and a handheld housing; Based on the above modules, the specific usage of the illuminance meter is as follows: S1: Equipment preparation and calibration: Assemble the intelligent photosynthetic illuminance meter, connect the sensor to the MCU, install the display and housing, and calibrate to ensure accurate data; S2: Set measurement parameters: After setting the measurement target, configure the sensor parameters to match the requirements of the crop type and growth stage; S3: Data collection: Bring a photosynthetic illuminance meter into the greenhouse, measure data at regular intervals, and store the data in the MCU or remote server; S4: Data preprocessing: clean the data to remove anomalies, standardize the sensor data, and ensure the quality of analysis; S5: Construct BP neural network model: determine the model structure according to the data dimension, randomly set the initial weight value and threshold, and prepare for BP network training; S6: Training BP neural network model: input preprocessed data, set training parameters, perform forward and backward propagation, iterate until the error converges, and complete model training; S7: Apply models and adjust planting strategies: Use BP models to predict yields, analyze data to adjust strategies, and optimize the environment to improve quality and yield.

[0005] Preferably, the equipment preparation and calibration in S1 refers to the need to ensure that the wavelength sensor, temperature sensor, hydrogen sensor, carbon dioxide sensor and oxygen sensor are accurately connected to the MCU microcontroller when assembling the intelligent photosynthetic illuminator. At the same time, the LED display and handheld housing should also be properly installed to facilitate real-time data monitoring and instrument operation. At the same time, before formal use, each sensor must undergo a strict calibration process to ensure the accuracy of its measurement data. Special attention should be paid to the wavelength sensor, whose calibration is particularly critical because it is necessary to accurately measure the illuminance of the chlorophyll absorption wavelength, which is an important indicator for evaluating the efficiency of photosynthesis.

[0006] Preferably, setting the measurement parameters in S2 means that before performing photosynthetic illumination measurement, it is necessary to clarify the measurement target, which includes the type of crops and their growth stage, because different crops have different requirements for light, temperature, gas concentration environmental factors. Clarifying the measurement target helps to adjust the measurement parameters according to actual needs. Subsequently, according to the specific requirements of the measurement target, it is necessary to carefully configure the measurement range, accuracy and parameters of each sensor.

[0007] Preferably, the specific steps of collecting data in S3 are as follows: Step 1: On-site measurement: Bring the intelligent photosynthetic illuminance meter into the greenhouse, and according to the predetermined measurement layout, go to each measurement point one by one. At each measurement point, start the instrument to collect data at a predetermined time interval, including the illumination of the chlorophyll absorption wavelength, temperature, carbon dioxide concentration, hydrogen concentration, and oxygen concentration. Step 2: Data recording: After the instrument completes data collection, ensure that all measured data is accurately recorded in the MCU microcontroller. If the remote data transmission function is set, the data will be automatically uploaded to the remote server. Step 3: Data storage and checking: After completing data collection at all measurement points, check the data records on the MCU microcontroller or remote server.

[0008] Preferably, the data preprocessing in S4 refers to the data cleaning and standardization work after completing the field measurement and data recording. It is necessary to carefully check all the collected data, identify and delete those abnormal values ​​or invalid data, which are caused by instrument failure and environmental interference. Deleting errors ensures the data quality of subsequent analysis. Then, data standardization processing is performed to convert the data of each sensor into a form with the same dimension and range. This is because the measurement values ​​of different sensors often have different units and orders of magnitude. These differences are eliminated through standardization processing.

[0009] Preferably, the specific steps of constructing the BP neural network model in S5 are as follows: Step 1: Determine the model structure: According to the dimension of the input data and the output target, design the architecture of the BP neural network model, including determining the number of neurons in the input layer, hidden layer and output layer. The number of neurons in the input layer is consistent with the dimension of the input data, the number of neurons in the output layer matches the number of output targets, and the number of hidden layers and the number of neurons in each layer are determined based on experience or experiments; Step 2: Randomly initialize connection weights: In a BP neural network, each neuron has a connection weight with other neurons. The weight determines the transmission strength of the signal in the network. After the model is built, these connection weights need to be randomly initialized. Usually, a small random number is used for initialization to ensure that the network training process can proceed smoothly. Step 3: Set and initialize the threshold: In addition to the connection weight, each neuron also has a threshold that determines whether the neuron is activated. Similar to the connection weight, the threshold also needs to be initialized, usually using a smaller random number.

[0010] Preferably, the specific steps of training the BP neural network model in S6 are as follows: Step 1: Input training data: Input the pre-processed data set as training data into the BP neural network model. The data is used to adjust the connection weights and thresholds of the neural network so that it can accurately predict the output target. Step 2: Set training parameters: Before starting training, you need to set a series of training parameters, including learning rate, number of iterations, and error convergence criteria. The learning rate determines the step size when updating the connection weights and thresholds each time, the number of iterations limits the maximum duration of the training process, and the error convergence criteria is used to determine whether the training has achieved the predetermined goal. Step 3: Training model: The training data is input into the neural network through the forward propagation algorithm, and the error between the predicted value and the actual value of the output layer is calculated. Then, the back propagation algorithm is used to adjust the connection weights and thresholds of the neural network layer by layer according to the error signal to reduce the prediction error. This process will be repeated until the predetermined training goal is reached or the maximum number of iterations is reached.

[0011] Preferably, the specific steps of applying the model and adjusting the planting strategy in S7 are as follows: Step 1: Predicting yield or sugar / carbohydrate content: Using the trained BP neural network model, taking the real-time collected greenhouse data as input, predict the crop yield or sugar / carbohydrate content C5; Step 2: Analyze the growth status of crops: Compare the predicted results with the actual situation and analyze the growth status of crops. If the predicted results show that the yield or sugar / carbohydrate content is lower than the expected value, combine the real-time data to determine whether there is a problem of insufficient light or low carbon dioxide concentration in the growth environment; Step 3: Adjust planting strategies: Based on the analysis results, make targeted adjustments to the light, temperature, humidity, and carbon dioxide concentration parameters in the greenhouse.

[0012] The beneficial effects of the present invention are as follows: The present invention combines high-precision sensors with advanced data analysis technology through an intelligent photosynthetic illuminance meter, which brings significant benefits to agricultural production. By accurately measuring the illumination, temperature, and gas concentration key parameters of the chlorophyll absorption wavelength, the instrument can reflect the growth environment of crops in real time. Its data preprocessing function ensures the analysis quality, and the construction and training of the BP neural network model further improves the accuracy of predicting crop yields or sugar / carbohydrate content. Users can adjust planting strategies in time according to the prediction results, optimize light and temperature environmental factors, thereby improving the growth conditions of crops and increasing yield and quality. In addition, the intelligent photosynthetic illuminance meter also has a remote data transmission function, which realizes real-time sharing and remote monitoring of data, and provides a more convenient and efficient management method for agricultural production. In summary, the application of the intelligent photosynthetic illuminance meter will help improve the intelligence and refinement level of agricultural production and promote the sustainable development of agriculture. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is a schematic diagram of the overall process of the present invention; Figure 2 This is a schematic diagram of the structure of the wavelength sensor of the present invention; Figure 3 This is a schematic diagram of the chlorophyll absorption wavelength structure of the present invention; Figure 4 Schematic diagram of the three-layer neuron model of the present invention; Figure 5 Schematic diagram of the neuron model of the present invention. DETAILED DESCRIPTION

[0014] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0015] like Figures 1 to 5 As shown, an embodiment of the present invention provides an intelligent photosynthetic illuminance meter, the illuminance meter package includes a multi-sensor module, an MCU microcontroller module, a BP algorithm center module, an LED display module and a handheld housing; Based on the above modules, the specific usage of the illuminance meter is as follows: S1: Equipment preparation and calibration: Assemble the intelligent photosynthetic illuminance meter, connect the sensor to the MCU, install the display and housing, and calibrate to ensure accurate data; S2: Set measurement parameters: After setting the measurement target, configure the sensor parameters to match the requirements of the crop type and growth stage; S3: Data collection: Bring a photosynthetic illuminance meter into the greenhouse, measure data at regular intervals, and store the data in the MCU or remote server; S4: Data preprocessing: clean the data to remove anomalies, standardize the sensor data, and ensure the quality of analysis; S5: Construct BP neural network model: determine the model structure according to the data dimension, randomly set the initial weight value and threshold, and prepare for BP network training; S6: Training BP neural network model: input preprocessed data, set training parameters, perform forward and backward propagation, iterate until the error converges, and complete model training; S7: Apply models and adjust planting strategies: Use BP models to predict yields, analyze data to adjust strategies, and optimize the environment to improve quality and yield.

[0016] Among them, the equipment preparation and calibration in S1 refers to the need to ensure that the wavelength sensor, temperature sensor, hydrogen sensor, carbon dioxide sensor and oxygen sensor are accurately connected to the MCU microcontroller when assembling the intelligent photosynthetic illuminator. At the same time, the LED display and handheld housing should also be properly installed to facilitate real-time data monitoring and instrument operation. At the same time, before formal use, each sensor must undergo a strict calibration process to ensure the accuracy of its measurement data. Special attention should be paid to the wavelength sensor, whose calibration is particularly critical because it is necessary to accurately measure the illuminance of the chlorophyll absorption wavelength, which is an important indicator for evaluating photosynthesis efficiency. Through fine assembly and calibration, the intelligent photosynthetic illuminator can provide reliable data support for agricultural production.

[0017] In S1, equipment preparation and calibration ensure the accuracy and reliability of the intelligent photosynthetic illuminance meter. Through fine assembly, each sensor is correctly connected to the MCU microcontroller to achieve real-time data monitoring and instrument operation. The strict calibration process, especially the precise calibration of the wavelength sensor, ensures the accuracy of the measured data and provides key indicators for evaluating photosynthesis efficiency. This step not only improves the performance of the instrument, but also provides accurate data support for agricultural production, which helps to optimize the growth environment of crops and improve yield and quality. Fine assembly and calibration lay a solid foundation for the widespread application of intelligent photosynthetic illuminance meters.

[0018] The setting of measurement parameters in S2 means that before measuring photosynthetic illumination, it is necessary to clarify the measurement target, which includes the type of crop and its growth stage. Because different crops have different requirements for environmental factors such as light, temperature, and gas concentration, clarifying the measurement target helps to adjust the measurement parameters according to actual needs. Subsequently, according to the specific requirements of the measurement target, it is necessary to carefully configure the measurement range, accuracy, and parameters of each sensor. For example, for crops with high chlorophyll content, it may be necessary to adjust the measurement range of the wavelength sensor to more accurately capture the illumination of the chlorophyll absorption wavelength. By accurately setting the measurement target and configuring the sensor parameters, it is ensured that the measurement results of the photosynthetic illumination meter are more in line with actual needs, providing more effective data support for agricultural production. The specific steps of collecting data in S3 are as follows: Step 1: On-site measurement: Bring the intelligent photosynthetic illuminance meter into the greenhouse, and according to the predetermined measurement layout, go to each measurement point one by one. At each measurement point, start the instrument to collect data at a predetermined time interval, including the illumination of the chlorophyll absorption wavelength, temperature, carbon dioxide concentration, hydrogen concentration, and oxygen concentration. Step 2: Data recording: After the instrument completes data collection, ensure that all measured data is accurately recorded in the MCU microcontroller. If the remote data transmission function is set, the data will be automatically uploaded to the remote server to achieve real-time data sharing and remote monitoring. Step 3: Data storage and inspection: After completing data collection at all measurement points, check the data records on the MCU microcontroller or remote server; ensure that all data has been completely saved and can be used for subsequent data analysis and agricultural production decisions.

[0019] In S3, the data collection process is rigorous and efficient. Through on-site measurements, key data in the greenhouse can be accurately captured, such as illumination at the wavelength absorbed by chlorophyll, temperature, carbon dioxide concentration, etc., to provide real-time information for agricultural production. The data recording link ensures that all data is accurately recorded and uploaded to the remote server to achieve real-time sharing and remote monitoring. Finally, the data storage and inspection steps ensure the integrity and availability of the data, providing a solid foundation for subsequent data analysis and decision-making. This process helps to improve the intelligence and refinement of agricultural production and provide strong support for the optimization of the crop growth environment.

[0020] The data preprocessing in S4 refers to the data cleaning and standardization work after completing the field measurement and data recording. It is necessary to carefully check all the collected data, identify and delete those abnormal values ​​or invalid data, which are caused by instrument failure and environmental factors. Deleting errors ensures the data quality of subsequent analysis. Then, data standardization is performed to convert the data of each sensor into a form with the same dimension and range. This is because the measurement values ​​of different sensors often have different units and orders of magnitude. Through standardization, these differences are eliminated. All data can be compared and analyzed on the same scale, which facilitates subsequent data mining and decision-making.

[0021] Data preprocessing plays a vital role in S4. By carefully checking and cleaning the data, we are able to identify and remove outliers and invalid data caused by instrument failure or environmental interference, thereby ensuring the data quality for subsequent analysis. In addition, data standardization converts the measurements of each sensor into a form with the same dimension and range, eliminating the unit and order of magnitude differences between different sensors. This step enables all data to be compared and analyzed on the same scale, which greatly facilitates subsequent data mining and decision-making, and helps to improve the accuracy and efficiency of crop growth environment monitoring.

[0022] The specific steps of constructing the BP neural network model in S5 are as follows: Step 1: Determine the model structure: According to the dimension of the input data (such as the illumination of the chlorophyll absorption wavelength, temperature, carbon dioxide concentration, etc.) and the output target (such as the yield of crops or the sugar / carbohydrate content C5), design the architecture of the BP neural network model, including determining the number of neurons in the input layer, hidden layer and output layer. The number of neurons in the input layer is consistent with the dimension of the input data, and the number of neurons in the output layer matches the number of output targets. The number of hidden layers and the number of neurons in each layer are determined based on experience or experiments; Step 2: Randomly initialize connection weights: In a BP neural network, each neuron has connection weights with other neurons. These weights determine the transmission strength of the signal in the network. After the model is built, these connection weights need to be randomly initialized. Usually, a small random number is used for initialization to ensure that the network training process can proceed smoothly. Step 3: Set and initialize the threshold: In addition to the connection weights, each neuron also has a threshold (also called bias) that determines whether the neuron is activated. Similar to the connection weights, the threshold also needs to be initialized, usually using a smaller random number.

[0023] When building a BP neural network model, it is crucial to clarify the model structure, randomly initialize the connection weights, and set the threshold. First, determine the model architecture based on the input data and output targets to ensure that the model matches the actual problem. Secondly, randomly initialize the connection weights to provide a starting point for model training and avoid falling into local optimality. At the same time, initialize the threshold to ensure that neurons are activated under appropriate conditions and improve the model's expressiveness. These steps not only lay a solid foundation for model training, but also enhance the adaptability and generalization ability of the model, which helps to improve the prediction accuracy of crop yields or sugar / carbohydrate content, and provide strong support for agricultural production.

[0024] The specific steps of training the BP neural network model in S6 are as follows: Step 1: Input training data: The preprocessed (such as data cleaning and standardization) data set is input into the BP neural network model as training data. The data is used to adjust the connection weights and thresholds of the neural network so that it can accurately predict the output target.

[0025] Step 2: Set training parameters: Before starting training, you need to set a series of training parameters, including learning rate, number of iterations (or training cycle), and error convergence criterion. The learning rate determines the step size when updating the connection weights and thresholds each time, the number of iterations limits the maximum duration of the training process, and the error convergence criterion is used to determine whether the training has achieved the predetermined goal. Step 3: Training model: The training data is input into the neural network through the forward propagation algorithm, and the error between the predicted value and the actual value of the output layer is calculated. Then, the back propagation algorithm is used to adjust the connection weights and thresholds of the neural network layer by layer according to the error signal to reduce the prediction error. This process will be repeated until the predetermined training goal is achieved (such as the error converges below a certain threshold) or the maximum number of iterations is reached.

[0026] By inputting the preprocessed training data, the model can learn the inherent laws of the data and accurately predict the output target. Secondly, setting reasonable training parameters, such as learning rate, number of iterations, and error convergence criteria, ensures the stability and efficiency of the training process. Finally, through the continuous iteration of the forward propagation and backpropagation algorithms, the model can gradually optimize the connection weights and thresholds, reduce the prediction error, and reach the predetermined goal. This process not only improves the prediction accuracy of the model, but also provides strong support for subsequent decision-making, which helps to achieve more efficient agricultural production and resource utilization.

[0027] The specific steps of applying the model and adjusting the planting strategy in S7 are as follows: Step 1: Predicting yield or sugar / carbohydrate content: Using the trained BP neural network model, the real-time greenhouse data (such as illumination at the chlorophyll absorption wavelength, temperature, and carbon dioxide concentration) collected in real time are used as input to predict the yield of crops or the sugar / carbohydrate content C5; Step 2: Analyze crop growth conditions: Compare the predicted results with the actual situation and analyze the growth conditions of crops. If the predicted results show that the yield or sugar / carbohydrate content is lower than expected, combine the real-time data to determine whether there are growth environment problems such as insufficient light or low carbon dioxide concentration.

[0028] Step 3: Adjust the planting strategy: According to the analysis results, adjust the light, temperature, humidity, and carbon dioxide concentration parameters in the greenhouse in a targeted manner. For example, if the light is insufficient, the light intensity can be increased; if the carbon dioxide concentration is too low, the carbon dioxide supply can be increased. By optimizing the growth environment of crops, the yield and quality can be improved.

[0029] The light from the greenhouse film passes through the K9 glass lens, passes through the filter, acts on the photocell, and obtains electric current, which is the intensity of the wavelength absorbed by chlorophyll.

[0030] The application of BP neural network model to predict crop yield or sugar / carbohydrate content, and combined with real-time data analysis to adjust planting strategies, has brought significant benefits. First, through accurate prediction, it is possible to understand the growth status of crops in a timely manner and avoid the risk of reduced yield or quality. Secondly, combined with real-time data, it is possible to accurately judge growth environment problems, such as insufficient light or low carbon dioxide concentration, so as to adjust the environmental parameters in the greenhouse in a targeted manner. This not only improves the yield and quality of crops, but also optimizes resource utilization and reduces production costs. In short, this method realizes the intelligence and refinement of agricultural production, and provides strong support for the sustainable development of agriculture.

[0031] The multi-sensor package includes a wavelength sensor, a temperature sensor, a hydrogen sensor, a carbon dioxide sensor, and an oxygen sensor.

[0032] The BP algorithm is programmed in MATLAB as follows: function y_hat=standard_BP_predict(data,v,r,w,h) [m,n]=size(data); for i=1:m x=data(i,:); A=x*v; %Input layer->hidden layer, the weights of each hidden layer unit b=fc_sigmod(Ar); % Output of the activation function after the hidden layer B=b*w; %Hidden layer->output layer, the weights of each output layer unit y_hat(i,:)=fc_sigmod(Bh); %Output of the activation function after the output layer end function y=fc_sigmod(x) y=1. / (1+exp(-x)); end end It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0033] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent photosynthetic illumination meter, characterized in that: The illuminance meter package has multiple sensor modules, MCU microcontroller modules, BP algorithm center modules, LED display modules and handheld housing; Based on the above modules, the specific usage of the illuminance meter is as follows: S1: Equipment preparation and calibration: Assemble the intelligent photosynthetic illuminance meter, connect the sensor to the MCU, install the display and housing, and calibrate to ensure accurate data; S2: Set measurement parameters: After setting the measurement target, configure the sensor parameters to match the requirements of the crop type and growth stage; S3: Data collection: Bring a photosynthetic illumination meter into the greenhouse, measure data at regular intervals, and store the data in the MCU or remote server; S4: Data preprocessing: clean the data to remove anomalies, standardize the sensor data, and ensure the quality of analysis; S5: Construct BP neural network model: determine the model structure according to the data dimension, randomly set the initial weight value and threshold, and prepare for BP network training; S6: Training BP neural network model: input preprocessed data, set training parameters, perform forward and backward propagation, iterate until the error converges, and complete model training; S7: Apply models and adjust planting strategies: Use BP models to predict yields, analyze data to adjust strategies, and optimize the environment to improve quality and yield.

2. The intelligent photosynthetic illumination meter according to claim 1, characterized in that: The equipment preparation and calibration in S1 refers to the need to ensure that the wavelength sensor, temperature sensor, hydrogen sensor, carbon dioxide sensor and oxygen sensor are accurately connected to the MCU microcontroller when assembling the intelligent photosynthetic illuminance meter. At the same time, the LED display and handheld housing should also be properly installed to facilitate real-time data monitoring and instrument operation. At the same time, before formal use, each sensor must undergo a strict calibration process to ensure the accuracy of its measurement data. Special attention should be paid to the wavelength sensor, whose calibration is particularly critical because it is necessary to accurately measure the illuminance of the chlorophyll absorption wavelength, which is an important indicator for evaluating the efficiency of photosynthesis.

3. The intelligent photosynthetic illumination meter according to claim 1, characterized in that: Setting the measurement parameters in S2 means that before measuring photosynthetic illumination, it is necessary to clarify the measurement target, which includes the type of crops and their growth stage, because different crops have different requirements for environmental factors such as light, temperature, and gas concentration. Clarifying the measurement target helps to adjust the measurement parameters according to actual needs. Subsequently, according to the specific requirements of the measurement target, it is necessary to carefully configure the measurement range, accuracy, and parameters of each sensor.

4. The intelligent photosynthetic illumination meter according to claim 1, characterized in that: The specific steps of collecting data in S3 are as follows: Step 1: On-site measurement: Bring the intelligent photosynthetic illuminance meter into the greenhouse, and according to the predetermined measurement layout, go to each measurement point one by one. At each measurement point, start the instrument to collect data at a predetermined time interval, including the illumination of the chlorophyll absorption wavelength, temperature, carbon dioxide concentration, hydrogen concentration, and oxygen concentration. Step 2: Data recording: After the instrument completes data collection, ensure that all measured data is accurately recorded in the MCU microcontroller. If the remote data transmission function is set, the data will be automatically uploaded to the remote server. Step 3: Data storage and checking: After completing data collection at all measurement points, check the data records on the MCU microcontroller or remote server.

5. The intelligent photosynthetic illumination meter according to claim 1, characterized in that: The data preprocessing in S4 refers to the data cleaning and standardization work after completing the field measurement and data recording. It is necessary to carefully check all the collected data, identify and delete those abnormal values ​​or invalid data, which are caused by instrument failure and environmental interference. Deleting errors ensures the data quality of subsequent analysis. Then, data standardization is performed to convert the data of each sensor into a form with the same dimension and range. This is because the measurement values ​​of different sensors often have different units and orders of magnitude. These differences can be eliminated through standardization.

6. The intelligent photosynthetic illumination meter according to claim 1, characterized in that: The specific steps of constructing the BP neural network model in S5 are as follows: Step 1: Determine the model structure: According to the dimension of the input data and the output target, design the architecture of the BP neural network model, including determining the number of neurons in the input layer, hidden layer and output layer. The number of neurons in the input layer is consistent with the dimension of the input data, the number of neurons in the output layer matches the number of output targets, and the number of hidden layers and the number of neurons in each layer are determined based on experience or experiments; Step 2: Randomly initialize connection weights: In a BP neural network, each neuron has a connection weight with other neurons. The weight determines the transmission strength of the signal in the network. After the model is built, these connection weights need to be randomly initialized. Usually, a small random number is used for initialization to ensure that the network training process can proceed smoothly. Step 3: Set and initialize the threshold: In addition to the connection weight, each neuron also has a threshold that determines whether the neuron is activated. Similar to the connection weight, the threshold also needs to be initialized, usually using a smaller random number.

7. The intelligent photosynthetic illumination meter according to claim 1, characterized in that: The specific steps of training the BP neural network model in S6 are as follows: Step 1: Input training data: Input the pre-processed data set as training data into the BP neural network model. The data is used to adjust the connection weights and thresholds of the neural network so that it can accurately predict the output target. Step 2: Set training parameters: Before starting training, you need to set a series of training parameters, including learning rate, number of iterations, and error convergence criteria. The learning rate determines the step size when updating the connection weights and thresholds each time, the number of iterations limits the maximum duration of the training process, and the error convergence criteria is used to determine whether the training has achieved the predetermined goal. Step 3: Training model: The training data is input into the neural network through the forward propagation algorithm, and the error between the predicted value and the actual value of the output layer is calculated. Then, the back propagation algorithm is used to adjust the connection weights and thresholds of the neural network layer by layer according to the error signal to reduce the prediction error. This process will be repeated until the predetermined training goal is reached or the maximum number of iterations is reached.

8. The intelligent photosynthetic illumination meter according to claim 1, characterized in that: The specific steps of applying the model and adjusting the planting strategy in S7 are as follows: Step 1: Predicting yield or sugar / carbohydrate content: Using the trained BP neural network model, taking the real-time collected greenhouse data as input, predict the crop yield or sugar / carbohydrate content C5; Step 2: Analyze the growth status of crops: Compare the predicted results with the actual situation and analyze the growth status of crops. If the predicted results show that the yield or sugar / carbohydrate content is lower than the expected value, combine the real-time data to determine whether there is a problem of insufficient light or low carbon dioxide concentration in the growth environment; Step 3: Adjust planting strategies: Based on the analysis results, make targeted adjustments to the light, temperature, humidity, and carbon dioxide concentration parameters in the greenhouse.

Images