Low-carbon energy-saving light supplementing method and device based on light requirement of plants

Abstract

The application provides a low-carbon and energy-saving light supplementing method and device based on light demand of plants, which comprises the following steps: obtaining environmental data of plants; inputting the environmental data into a light supplementing decision model, wherein the light supplementing decision model outputs a target value of light-related quantity; based on the target value, adjusting a light supplementing subsystem of the plants, so that a real-time value of the light-related quantity collected in the environment of the plants approaches the target value; wherein the light supplementing decision model can obtain the target value by optimizing a minimum value of a comprehensive weighted value of a negative value of predicted photosynthetic rate, net carbon emission of light supplementing and light supplementing electric power; and the application takes improving photosynthetic rate of plants, reducing light supplementing power and carbon emission as the target, calculates the negative value of predicted photosynthetic rate, net carbon emission of light supplementing and light supplementing electric power under the light supplementing scene, obtains the optimal light supplementing target value, and realizes the low-carbon and energy-saving light supplementing for the plants.

Description

Technical Field

[0001] This invention relates to the field of crop supplemental lighting control technology, and in particular to a low-carbon and energy-saving supplemental lighting method and device based on the light requirements of plants. Background Technology

[0002] Light provides the energy for plant photosynthesis and serves as an environmental signal regulating plant growth and development. Light conditions, including light quality, light intensity, and photoperiod, are among the important environmental variables regulating plant growth and development. Net photosynthetic rate (Pn) represents the accumulation of organic matter by a plant per unit time through photosynthesis and is an important indicator for evaluating the photosynthetic status of a plant. On the one hand, plant photosynthesis is significantly affected by the environment, especially photosynthetic parameters that characterize the plant's light requirements, such as the light saturation point, light compensation point, and the differential (difference) value of the net photosynthetic rate. These parameters drift with changes in environmental factors such as air and soil temperature and humidity, and environmental carbon dioxide (CO2) concentration. Correspondingly, the cost of supplemental lighting varies depending on the environmental conditions at different times of the day. On the other hand, influenced by seasons and growth stages, plants themselves differ in morphology and structure. Plants at different growth stages exhibit distinct characteristics in terms of light requirements in terms of wavelength and bandwidth, and the benefits of supplemental lighting are also affected by fluctuations in yield and price at different times.

[0003] Wireless Sensor Networks (WSNs) are an important tool for agricultural environmental monitoring. Their low cost ensures a sufficient number of devices available for use in the environment, and the measurements can fully reflect the spatiotemporal variability of the current environment. Currently, WSNs are widely used for the automatic monitoring of greenhouse environmental parameters. In recent years, the global market value of LED (Light Emitting Diode) plant lighting applications has grown rapidly. Compared with traditional supplemental lighting such as high-pressure sodium lamps, LED plant supplemental lighting has more precise and controllable wavelengths and bandwidths, allowing it to match the light requirements of plants and enabling high-efficiency plant supplemental lighting. Simultaneously, with the development of LED packaging and integration technology, its luminous efficiency is continuously improving, and the heat dissipation of the lamps is constantly decreasing, resulting in lower operating costs for plant supplemental lighting. Machine learning is a core technology in the fields of artificial intelligence and pattern recognition. Its theories and methods are widely used to solve complex problems in engineering and scientific fields. Machine learning algorithms such as Support Vector Machine (SVM), Single Hidden Layer Feedforward Neural Network (SLFN), and Random Forest (RF) provide technical support for the establishment of plant light requirement models and supplemental lighting decision models, playing an important role in plant management and yield prediction. Artificial intelligence algorithms such as Evolutionary Computation (EC) also provide a foundation for decision-making under multi-constraint or multi-objective regulation. Modeling the underlying mechanisms is the foundation for information-based and automated management in controlled environments such as greenhouses. The establishment of models for photosynthesis prediction and light environment decision-making, with machine learning and evolutionary computing as the main supporting technologies, is key to achieving low-carbon and energy-saving control. These models complement traditional mathematical models in terms of generalization and functionality. Based on plant models, the calculation of target values ​​at the decision level, considering multiple constraints such as energy consumption, will provide core technical support for environmental control systems such as supplemental lighting.

[0004] In addition, supplemental lighting in greenhouses can increase the net photosynthetic rate of plants, enabling high-quality and high-yield plants in controlled environments such as greenhouses. However, supplemental lighting also brings additional carbon emissions. High-power, high-intensity supplemental lighting is not necessarily suitable for the light requirements of plants under all environmental conditions, thus increasing production costs.

[0005] The information disclosed in this background section is intended only to enhance the understanding of the general background of the invention and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0006] To address the problems existing in the prior art, this invention provides a low-carbon, energy-saving supplemental lighting method and device based on the light requirements of plants.

[0007] The present invention provides a low-carbon, energy-saving supplemental lighting method based on the light requirements of plants, the method comprising:

[0008] Obtain environmental data of the plants;

[0009] The environmental data is input into the supplemental lighting decision model, and the supplemental lighting decision model outputs the target value of the illumination-related quantity.

[0010] Based on the target value, the supplemental light system of the plant is adjusted so that the real-time values ​​of light-related quantities collected in the environment where the plant is located approach the target value.

[0011] in,

[0012] The supplemental lighting decision model can optimize and obtain the target value by minimizing the combined weighted value of the predicted negative photosynthetic rate, the net carbon emissions from supplemental lighting, and the photoelectric power from supplemental lighting.

[0013] Optionally, the method for obtaining the predicted photosynthetic rate further includes:

[0014] Obtain the target value;

[0015] The target value is input into the LS-SVR model, and the LS-SVR model outputs the predicted photosynthetic rate corresponding to the target value;

[0016] in,

[0017] The kernel function of the LS-SVR model uses radial basis functions, and the formula for the radial basis functions is as follows:

[0018]

[0019] Where, x i and x j Corresponding to the two samples of the LS-SVR model, ||x i -x j ||For x i and x j The Euclidean distance, where σ is the extension constant of the radial basis functions.

[0020] Optionally, the hyperparameters of the LS-SVR model include the regularization coefficient γ and the expansion constant σ of the radial basis functions;

[0021] The training method for the LS-SVR model further includes:

[0022] Obtain training samples;

[0023] Based on the samples, a test set, a validation set, and a training set are formed through cross-validation.

[0024] The individual position in the evolutionary computation is encoded as a hyperparameter to be optimized; the average value of the photosynthetic rate prediction error of the hyperparameter represented by the individual position on the validation set is obtained, and the average value is used as the fitness value corresponding to the individual; the individual position and the fitness value corresponding to the individual are updated until the maximum number of iterations is reached or the error is less than a set value, and the optimal hyperparameter is obtained.

[0025] Based on the optimal hyperparameters, the LS-SVR model is trained using the training set.

[0026] Optionally, obtain training samples, including:

[0027] The photon flux density and the proportion of several different wavelengths of light in the environment where the plant is located are obtained by a photon sensor and used as the input of the LS-SVR model.

[0028] The actual photosynthetic rate of the plant corresponding to the light quantum flux density and the light wave ratio is obtained by a photosynthetic rate meter and used as the output of the LS-SVR model.

[0029] The input quantity and the corresponding output quantity are used as a sample of the LS-SVR model.

[0030] Optionally, the method for optimizing to obtain the target value further includes at least one of the first constraint and the second constraint;

[0031] The first constraint includes:

[0032] The current of the supplementary light in the supplementary lighting subsystem is less than or equal to the maximum rated current;

[0033] The second constraint includes:

[0034] The target value is not less than the corresponding value of the light-related quantity of the plant at the light compensation point;

[0035] The method for obtaining the corresponding values ​​of light-related quantities of the plant at the light compensation point further includes:

[0036] Based on the trained LS-SVR model, the corresponding values ​​of the illumination correlation quantities at the light compensation points are obtained by the binary search method.

[0037] Optionally, the method for obtaining the net carbon emissions from supplemental lighting further includes:

[0038] Obtain the carbon emissions corresponding to the energy consumed by the supplementary lighting subsystem during operation;

[0039] The amount of carbon fixed by the plant due to supplemental lighting is the amount of carbon fixed by the plant after supplemental lighting under the same conditions minus the amount of carbon fixed by the plant before supplemental lighting.

[0040] The net carbon emissions from supplemental lighting are calculated by subtracting the amount of carbon fixed by supplemental lighting from the corresponding carbon emissions from the energy consumed.

[0041] Optionally, the supplementary lighting in the supplementary lighting subsystem is controlled by a PWM signal;

[0042] Based on the target value, the supplemental photonics system of the plant is adjusted so that the real-time values ​​of light-related quantities collected in the environment where the plant is located approach the target value, including:

[0043] The PWM signal is controlled using an incremental proportional-integral-derivative algorithm.

[0044] The technical solution of the present invention also provides a low-carbon and energy-saving supplemental lighting device based on the light requirements of plants, the device comprising:

[0045] The acquisition module is used to acquire environmental data of the plants;

[0046] The decision module is used to input the environmental data into the supplemental lighting decision model, and the supplemental lighting decision model outputs the target value of the illumination-related quantity.

[0047] An adjustment module is used to adjust the supplemental lighting system of the plant based on the target value, so that the real-time values ​​of light-related quantities collected in the environment where the plant is located approach the target value.

[0048] in,

[0049] The supplemental lighting decision model can optimize and obtain the target value by minimizing the combined weighted value of the predicted negative photosynthetic rate, the net carbon emissions from supplemental lighting, and the photoelectric power from supplemental lighting.

[0050] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps of the low-carbon energy-saving supplemental lighting method based on the light requirements of plants as described in any of the above claims.

[0051] The present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, wherein the computer program, when executed by a processor, implements the steps of the low-carbon energy-saving supplemental lighting method based on the light requirements of plants as described in any of the above claims.

[0052] The present invention provides a low-carbon and energy-saving supplemental lighting method and device based on the light requirements of plants. With the goal of improving the photosynthetic rate of plants and reducing the supplemental lighting power and carbon emissions, it calculates the negative value of the predicted photosynthetic rate, the net carbon emissions of supplemental lighting, and the supplemental lighting power under the supplemental lighting scenario to obtain the optimal supplemental lighting target value, thereby achieving low-carbon and energy-saving supplemental lighting for plants. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0054] Figure 1 This is a flowchart illustrating a low-carbon, energy-saving supplemental lighting method based on plant light requirements, according to an embodiment of the present invention.

[0055] Figure 2 This is an architectural diagram of a low-carbon, energy-saving supplemental lighting system based on plant light requirements, according to an embodiment of the present invention.

[0056] Figure 3 This is a schematic diagram illustrating the supplemental lighting decision-making process according to an embodiment of the present invention;

[0057] Figure 4 This is a schematic diagram of the optimal solution set of the supplementary lighting target value according to an embodiment of the present invention;

[0058] Figure 5 This is an overall flowchart of the supplementary lighting subsystem according to an embodiment of the present invention;

[0059] Figure 6 This is a flowchart of the PID control for supplemental lighting according to an embodiment of the present invention;

[0060] Figure 7 This is a schematic diagram of a low-carbon, energy-saving supplemental lighting device based on plant light requirements, according to an embodiment of the present invention.

[0061] Figure 8 This is a schematic diagram of the physical structure of an electronic device provided by the present invention. Detailed Implementation

[0062] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0063] The low-carbon, energy-saving supplemental lighting method based on plant light requirements provided in this application will be described in detail below with reference to the accompanying drawings, through specific embodiments and application scenarios.

[0064] Example 1

[0065] Figure 1 This is a flowchart illustrating a low-carbon, energy-saving supplemental lighting method based on plant light requirements, as described in an embodiment of the present invention. Figure 1 As shown, the technical solution of the present invention provides a low-carbon and energy-saving supplemental lighting method based on the light requirements of plants, the method comprising the following steps:

[0066] S100, Obtain environmental data of the plants;

[0067] S200. Input environmental data into the supplemental lighting decision model, and the supplemental lighting decision model outputs the target value of light-related quantities.

[0068] S300. Based on the target value, adjust the plant's supplemental light system so that the real-time values ​​of light-related quantities collected in the plant's environment approach the target value.

[0069] in,

[0070] The supplemental lighting decision model can optimize and obtain the target value by minimizing the combined weighted value of the predicted negative value of photosynthetic rate, net carbon emissions from supplemental lighting, and supplemental photoelectric power.

[0071] In this embodiment, the method for obtaining the predicted photosynthetic rate further includes:

[0072] Get the target value;

[0073] Input the target value into the LS-SVR model, and the LS-SVR model outputs the predicted photosynthetic rate corresponding to the target value;

[0074] The kernel function of the LS-SVR model uses radial basis functions, and the formula for the radial basis functions is as follows:

[0075]

[0076] Where, x i and x j Corresponding to the two samples of the LS-SVR model, ||x i -x j ||For x i and x j The Euclidean distance, where σ is the extension constant of the radial basis functions.

[0077] In this embodiment, the hyperparameters of the LS-SVR model include the regularization coefficient γ and the extension constant σ of the radial basis functions;

[0078] The training methods for the LS-SVR model further include:

[0079] Obtain training samples;

[0080] Based on the samples, cross-validation is used to form a test set, a validation set, and a training set;

[0081] Encode the individual position in the evolutionary computation as a hyperparameter to be optimized; obtain the average value of the photosynthetic rate prediction error of the hyperparameter represented by the individual position on the validation set, and use the average value as the fitness value of the corresponding individual; update the individual position and the corresponding fitness value until the maximum number of iterations is reached or the error is less than the set value, and obtain the optimal hyperparameter;

[0082] The LS-SVR model is trained using the training set based on the optimal hyperparameters.

[0083] In this embodiment, obtaining training samples includes:

[0084] The photon flux density in the environment where the plant is located and the proportion of several different wavelengths of light are obtained by photon sensors and used as input to the LS-SVR model.

[0085] The actual photosynthetic rate of the plant corresponding to the photon flux density and light wave ratio is obtained by the photosynthetic rate meter and used as the output of the LS-SVR model.

[0086] The input and the corresponding output are used as a sample of the LS-SVR model.

[0087] In this embodiment, the method for optimizing the acquisition of the target value further includes at least one of the first constraint and the second constraint;

[0088] The first constraint includes:

[0089] In the supplementary lighting system, the current of the supplementary lamp is less than or equal to the maximum rated current;

[0090] The second constraint includes:

[0091] The target value should not be less than the corresponding value of the light-related quantity of the plant at the light compensation point;

[0092] The method for obtaining the corresponding values ​​of light-related quantities of plants at the light compensation point further includes:

[0093] Based on the trained LS-SVR model, the corresponding values ​​of illumination correlation quantities at the light compensation point are obtained by the binary method.

[0094] In this embodiment, the method for obtaining the net carbon emissions from supplemental lighting further includes:

[0095] Obtain the carbon emissions corresponding to the energy consumed by the supplementary lighting subsystem during operation;

[0096] The amount of carbon fixed by the plant due to supplemental lighting is the amount of carbon fixed by the plant after supplemental lighting under the same conditions minus the amount of carbon fixed by the plant before supplemental lighting.

[0097] The net carbon emissions from supplemental lighting are calculated by subtracting the amount of carbon sequestrated by supplemental lighting from the corresponding carbon emissions from energy consumption.

[0098] In this embodiment, the fill light in the fill light subsystem is controlled by a PWM signal;

[0099] Based on the target value, adjust the plant's supplemental photon system so that the real-time values ​​of light-related quantities collected in the plant's environment approach the target value, including:

[0100] The incremental proportional-integral-derivative algorithm is used to control the PWM signal.

[0101] This embodiment aims to improve the photosynthetic rate of plants, reduce supplemental lighting power and carbon emissions. It calculates the negative value of the predicted photosynthetic rate, the net carbon emissions from supplemental lighting, and the photoelectric power of supplemental lighting under the supplemental lighting scenario, and obtains the optimal supplemental lighting target value to achieve low-carbon and energy-saving supplemental lighting for plants.

[0102] Example 2

[0103] Figure 2 This is an architectural diagram of a low-carbon, energy-saving supplemental lighting system based on plant light requirements, as described in an embodiment of the present invention. Figure 2 As shown, it is mainly divided into four parts: application layer, network layer, environment perception layer and execution layer (also power perception). Among them, the decision model is embedded in the remote management platform through software. Figure 3 This is a schematic diagram of the supplemental lighting decision-making process according to an embodiment of the present invention, such as... Figure 3 As shown, the supplemental lighting decision in this embodiment is disclosed. The photosynthesis prediction model in the figure is the LS-SVR model, and the decision model in the figure is the supplemental lighting decision model.

[0104] This embodiment utilizes WSN technology to achieve real-time monitoring of greenhouse environment and supplemental lighting power. A WSN monitoring system based on a System-on-Chip (SoC) (CC2530) is used to complete real-time data acquisition. A coordinator, router, and sensor nodes form the underlying wireless network transmission system based on the ZigBee protocol, enabling wireless data transmission. The network topology is tree-like. The coordinator sends data via serial port to a Data Transfer Unit (DTU) based on a 5G mobile information system. The DTU and server communicate using the Message Queuing Telemetry Transport (MQTT) protocol. Sensor nodes are connected to air temperature and humidity (SHT31), soil temperature and humidity, carbon dioxide (CO2) concentration, and (different wavelengths) photon sensors. The acquisition period for light intensity and air temperature and humidity is set to 10 minutes, ensuring high data continuity and reliability. This provides an effective and reliable research platform and important data support for studying the spatiotemporal distribution differences of greenhouse light intensity, air temperature, and humidity. Meanwhile, the end effector, i.e., the supplementary light, is connected to the AC / DC power metering module (IM1253B) to realize real-time measurement of supplementary light power. The feedback from this module is transmitted to the microcontroller (ESP32) via a transistor-to-transistor logic (TTL) serial port. The ESP32 reads the actual supplementary light power P and transmits the power data to the access point (AP) via the User Datagram Protocol (UDP). The AP then packages the power data and transmits it to the RS-232 to TTL module. Finally, the DTU uses the MQTT protocol to feed the data back to the cloud for power consumption calculation, thereby providing support for the calculation of the supplementary light decision model.

[0105] To elucidate the response mechanism of plant photosynthesis to the environment, especially light, and to provide plant information support for calculating plant light requirements and deciding on supplemental lighting targets, this embodiment uses Least Square Support Vector Regression (LS-SVR) to adjust the model hyperparameters based on the features of the collected dataset, thereby obtaining the optimal photosynthetic rate prediction model.

[0106] This embodiment includes light quality regulation. Taking red and blue light as examples, the model input includes photon flux density (PPFD), red light (wavelength 642nm~682nm) proportion (R), blue light (wavelength 428nm~464nm) proportion (B), air temperature (Ta), relative humidity (RH), soil temperature (Ts), relative soil moisture content (M), and CO2 concentration. The output is the real-time Pn of the plant. The range and gradient of the input are shown in Table 1. The environmental parameter values ​​in the incubator are set according to the orthogonal experiment, and the actual environmental parameter values ​​are obtained using WSN. The actual values ​​are used as the input of the modeling dataset. At the same time, the Pn of the plant under the current environmental conditions is measured using a photosynthetic rate meter (LI-6400XT / LI-6800, LI-COR, USA), and it is used as the output corresponding to the input.

[0107]

[0108]

[0109] Table 1

[0110] The input data is preprocessed raw data, including outlier removal using box plots and data standardization based on minimum and maximum values. The LS-SVR model avoids the computational problem of mapping samples from the original space to a high-dimensional space through a kernel function, so the input data does not need to be dimensionality reduced, thus preventing a decrease in prediction accuracy. The kernel function is chosen as the radial basis function (RBF), as shown in Equation (1).

[0111]

[0112] Where, x i and x j Corresponding to the two samples of the LS-SVR model, ||x i -x j ||For x i and x j The Euclidean distance, where σ is the extension constant of the radial basis functions;

[0113] To obtain an accurate photosynthetic rate prediction model, the hyperparameters of the LS-SVR model were optimized using a combination of cross-validation (CV) and evolutionary computation (CV-EC). The model's hyperparameters include the regularization coefficient γ and the RBF extension constant σ, which determine the model's minimum fitness error and the kernel function's bandwidth, respectively, thus influencing the model's smoothness and function selectivity, and significantly impacting the accuracy of Pn prediction. By default, the hyperparameters fail to reflect the characteristics of environmental parameters and photosynthetic data, leading to bias in the selection of influential samples.

[0114] The steps for optimizing model hyperparameters using CV-EC and establishing a photosynthetic rate prediction model based on LS-SVR are as follows:

[0115] First, 20% of the samples are used as the final test set according to the sample size. The remaining n samples are divided using the leave-one-out method. That is, n datasets are obtained by copying, and each dataset includes the remaining n samples. Then, one unique sample is selected from each dataset as the validation set. In this way, each sample serves as a validation set once in these n datasets.

[0116] Then, the individual locations in the EC are encoded as hyperparameter values ​​to be optimized, and the fitness value corresponding to the individual is the average of the photosynthetic rate prediction error of the hyperparameter represented by the individual location on n validation sets. The individual locations and the corresponding fitness values ​​are updated until the maximum number of iterations is reached or the error is less than the set value, thus obtaining the optimal hyperparameters under the current environment and photosynthetic dataset.

[0117] Finally, using the optimal hyperparameters, the model was trained on 80% of the training set to obtain an accurate photosynthetic rate prediction model, the performance of which can be evaluated using the test set.

[0118] To obtain the minimum light requirement of plants, i.e., the light compensation point under different environmental conditions, and thus provide constraints for supplemental lighting, a photosynthetic rate prediction model is used as the objective function. A bisection method is employed to extract the PPFD value corresponding to an environment where Pn is 0, which is then used as the light compensation point and represented by matrix Q. The specific steps for extracting the plant light compensation point using the bisection method are as follows:

[0119] Let the functional expression of the photosynthetic rate prediction model be Pn(x). First, let p m To measure the maximum light intensity, p l To measure the minimum light intensity value (0), according to the characteristics of photosynthetic rate change, we must have Pn(p m If ) > 0 and Pn(0) < 0, then due to the continuity of the corresponding function in the prediction model, the light saturation point p0 lies between these two points. Then, let... And determine the sign of Pn(px). If Pn(p) x If ) > 0, then take p0 ∈ (p l p x Let P m =p x If Pn(p x If ) = 0, then take p0 = p x If Pn(p x If ) < 0, then take p0 ∈ (p x p m Let p l =p x This is considered one iteration; finally, repeat the iteration until p...m With p l The distance satisfies the set precision or makes Pn(p x If the value reaches the preset value, the iteration stops, and the corresponding p that meets the iteration termination condition is removed. m p l or p x The value is saved and output as the optical compensation point Q under this condition.

[0120] Greenhouse supplemental lighting increases plant Pn, but also brings additional carbon emissions and energy consumption. This embodiment calculates and defines supplemental lighting carbon emissions using equations (2-5). In general, the net carbon emissions from supplemental lighting are equal to the carbon emissions corresponding to the energy consumed by LED operation minus the amount of carbon fixed by the plant due to supplemental lighting, as shown in equation (2); the amount of carbon fixed by the plant due to supplemental lighting is equal to the amount of carbon fixed by the plant after supplemental lighting minus the amount of carbon fixed by the plant before supplemental lighting, as shown in equation (3).

[0121] C SL =C LED -ΔC P (2)

[0122] ΔC P =C P2 -C P1 (3)

[0123] Among them, C SL To compensate for net carbon emissions from lighting, C LED The carbon emissions from the operation of supplemental lighting, ΔC P To enhance carbon fixation in plants through supplemental lighting, C P2 C represents the carbon sequestration capacity of the plant after supplemental lighting. P1 The carbon sequestration of plants without supplemental light (under natural conditions) is given above, and all variables are in grams (g).

[0124] Specifically, the carbon emissions from supplemental lighting are mainly determined by the energy consumed by the LED lamp and the light efficiency of the lamp itself, as shown in Equation (4); on the other hand, the gain in carbon fixation of the plant after supplemental lighting is determined by the plant's Pn under supplemental lighting and the leaf area of ​​the plant itself under supplemental lighting radiation, as shown in Equation (5).

[0125]

[0126]

[0127] Among them, PPFD LED This refers to the actual photonic flux density of the supplemental lighting, measured in micromoles per square meter per second (μmol·m²). -2 ·s -1 S represents the actual effective irradiance area of ​​the supplementary light, in square meters (m²). -2 PPFLED The photosynthetic photon flux efficiency of the supplemental lighting lamp is expressed in micromoles per second per watt (μmol·s). -1 ·W -1 );C a The carbon consumption corresponding to the electricity consumed by the current supplemental lighting is expressed in grams per kilowatt-hour (g·kWh). -1 Pn(E) represents the net photosynthetic rate of the plant under environmental condition E, expressed in micromoles of carbon dioxide per square meter per second (μmol CO2·m²). -2 ·s -1 E(x) is the environmental parameter value matrix at time x, and the environmental parameters include the variables in Table 1; the supplementary lighting time interval is [x1, x2], and the unit is seconds (s); S P The total leaf area of ​​the plant that has received effective radiation, expressed in square meters (m²). 2 ).

[0128] This embodiment integrates a multi-channel LED plant light with a PWM driver and an AC / DC power metering module. While controlling the supplementary light, it calculates the power of the current supplementary light as shown in equation (6) and uploads it through a forwarding module, thus realizing real-time power measurement feedback.

[0129]

[0130] Where P is the actual operating power of the fill light, in watts (W); U is the fill light voltage, in volts (V); and I is the fill light current, in amperes (A). The power factor.

[0131] To achieve cost-effective supplemental lighting, while increasing the plant's photosynthetic level, this embodiment abstracts the supplemental lighting target value problem into a constrained multi-objective optimization (MOO) problem, as shown in equation (7-9), striving to reach an optimal balance between increased yield and energy conservation. The target value t is a k-dimensional variable (u, v1, ..., v...). k-1 u represents the total target light intensity, measured in micromoles per square meter per second (μmol·m²). -2 ·s -1 v1, ..., v k-1These represent the percentage of each color of light (wavelength band), expressed as a percentage (%). In this embodiment, red-blue dual-color light supplementation is used as an example. Let the target value t correspond to the total target light intensity value u. Then, the total PPFD value of the red-blue dual-channel supplemental lighting is u, where the PPFD value of red light is u·v, and correspondingly, the PPFD value of blue light is (uu·v). Combining the aforementioned photosynthetic rate prediction model, with the net carbon emissions from supplemental lighting and the power of supplemental lighting as targets, and the maximum rated current of the supplemental lighting and the light requirement of the plant as constraints, the target light intensity value is optimized and calculated to obtain the optimal solution set.

[0132] min t F(t) = [Pn′(t), C SL (t), P(t)] (7)

[0133] Make t≥Q

[0134] I i (t)≤a i , i∈[1,k] (8)

[0135] Let A be the optimal solution set of MOO, then

[0136] A={t|t≥0,I i (t)≤m i ,i∈[1,k]} (9)

[0137] Where F(t) is the objective function; Pn′(t) is the negative of the net photosynthetic rate of the plant under the corresponding environmental conditions when the PPFD value is t, and the unit is micromoles per square meter per second (μmol·m²). -2 ·s -1 );C SL P(t) represents the carbon emissions of the luminaire at a PPFD value of t, in grams (g); P(t) represents the energy consumption of the luminaire at a PPFD value of t, in watts (W); Q represents the light compensation point under the corresponding environmental conditions, in micromoles per square meter per second (μmol·m²). -2 ·s -1 );I i (t) represents the actual current of the corresponding color supplemental light, a i The rated maximum current is expressed in amperes (A), and i is the circuit number of the supplementary light, corresponding to k different colors of light. In this embodiment, red and blue light are used as examples, so k = 2.

[0138] The above MMO solution can be obtained using methods such as the main linear weighting method, the objective method, and the approximation objective method. Taking the linear weighting method as an example, the sub-objective function Pn(t) is set according to its importance, and C... SL (t), the weights λ of P(t) l(1 = 1, 2, 3), thus transforming F(t) into a single-objective optimization problem, as shown in equation (10-11). For a given λ∈N, the optimal solution set of the above problem is an efficient solution to the MOO problem.

[0139] min t [λ1Pn′(t)+λ2C SL [(t)+λ3P(t)] (10)

[0140] Make t≥Q

[0141] I i (t)≤a i , i∈[1,k] (11)

[0142] in,

[0143] also, Figure 4 This is a schematic diagram of the optimal solution set of the supplementary lighting target value in an embodiment of the present invention, as shown below. Figure 4 As shown, EC algorithms such as Multi-Object Particle Swarm Optimization (MOPSO) and Multi-Objective Evolutionary Algorithms can be applied to solve the MOO problem. Taking MOPSO as an example, the problem-solving steps are as follows: First, initialize the positions and velocities of the particles to form an initial population, calculate the fitness value of the initial population, and copy the non-dominated solutions to the archive set; then, select based on the particle density information in the archive set, update the positions and velocities of the particles in the population, copy the non-dominated solutions of the new generation population to the archive set, and if the number of particles in the archive set exceeds a specified value, delete some particles from the archive set through a truncation operation; finally, stop the search when the maximum number of iterations, computational accuracy, or maximum stagnation steps of the optimal solution are reached, and output the optimal solution set.

[0144] The optimized supplemental lighting target values ​​and corresponding environmental conditions are saved to establish a supplemental lighting decision model, which is stored on the server. This model is triggered to adjust the lighting each time node data is transmitted to the Transmission Control Protocol (TCP) server created by Node.js (the runtime environment). The input to the supplemental lighting decision model is the environmental parameters detected by the node in real time, i.e., environmental variables other than illumination-related quantities such as PPFD, R, and B in Table 1. The output is illumination-related quantities such as PPFD, R, and B. Simultaneously, the corresponding inputs, outputs, and time points are stored in the database to store the target values, allowing for real-time viewing on both computer and mobile devices.

[0145] The system's end effector is a multi-channel supplementary light, meaning each supplementary light contains different channels corresponding to different light qualities. For example, in red-blue supplementary lighting, each supplementary light contains two channels: red and blue. The cloud-based decision layer sends control commands to the DTU via the MQTT protocol. Upon receiving the commands, the DTU packages the output of the supplementary lighting decision model from the decision layer into commands and sends them down via the RS-232 level protocol. At the RS-232 to TTL converter, the RS-232 level signal is converted to a TTL level signal. The ESP32 at the communication module receives the TTL signal via the serial port, processes the control commands, and then broadcasts the control commands down as an AP node using the UDP protocol. The instructions are shown in Table 2. The first bit is the start bit of the data, the second bit is the controlled device identification number (ID), which is accurate to the specific branch of the supplementary light, the third bit is the high 8 bits of the target light intensity output by the supplementary light decision model, the fourth bit is the low 8 bits of the target light intensity output by the supplementary light decision model (i.e., a total of 16 bits of the target light intensity output), the fifth bit is the high 8 bits of the photonic quantum sensor transmitted by the perception layer, the sixth bit is the low 8 bits of the photonic quantum sensor transmitted by the perception layer (i.e., a total of 16 bits of the current value of the photonic quantum sensor output), the seventh bit is the current time of the transmitted instructions, the eighth bit is the data check bit, and the ninth bit is the stop bit.

[0146] First Second place Third place Fourth place S NUM <![CDATA[TargetValue High ]]> <![CDATA[TargetValue Low ]]> Fifth Sixth Seventh Eighth place <![CDATA[CurrentValue High ]]> <![CDATA[CurrentValue Low ]]> Time CRC Ninth E

[0147] Table 2

[0148] Figure 5 This is an overall flowchart of the supplementary lighting subsystem according to an embodiment of the present invention, as follows: Figure 5 As shown, the device ID and target light intensity TargetValue (i.e., TargetValue) are obtained through algorithm parsing. High and TargetValue Low (a combination of the two light intensities) and the current light intensity CurrentValue (i.e., CurrentValue) High and CurrentValue Low (combination).

[0149] The supplementary light adjusts its brightness according to the pulse width modulation (PWM) signal. In order to establish the correspondence between the output light intensity and the PWM signal, this embodiment uses the incremental proportional-integral-derivative (PID) algorithm to control the PWM signal, so as to realize the illumination control based on the target light intensity. Figure 6 The flowchart of the supplementary lighting PID control in an embodiment of the present invention is as follows: Figure 6As shown, a corresponding PWM signal is generated through a timer, and the increment Δ of the PWM is calculated using a PID algorithm. PWM The corresponding signal is transmitted to the PWM controller of the lamp, thereby adjusting the ambient light intensity to continuously approach and reach the target value.

[0150] This embodiment, while meeting the basic light requirements of the plants, also offers advantages in reducing supplemental lighting energy consumption and carbon emissions during the production process, including:

[0151] (1) Based on the interaction between plants and the environment, environmental and photosynthetic data were collected using WSN and photosynthetic rate measurement instrument, respectively. A photosynthetic rate prediction model was established based on LS-SVR, and the parameters of the prediction model were optimized for the characteristics of photosynthetic data using the CV-EC method.

[0152] (2) By using the bisection method with the corresponding function of the photosynthetic rate prediction model as the object, light compensation points under different environmental conditions and plant growth stages were extracted to provide minimum light requirement guarantee for light intensity control.

[0153] (3) In the supplementary lighting subsystem, the AC / DC power calculation module is integrated with the LED lamp, and the data is uploaded to the server through the forwarding module to realize the power perception of the supplementary lighting;

[0154] (4) A method for calculating net carbon emissions under supplementary lighting scenarios was proposed, which provides a basis for optimizing supplementary lighting target values ​​under constraints;

[0155] (5) The target value of supplemental lighting was determined as an MMO problem and the optimal solution set was obtained through algorithms such as linear weighting method and MOPSO. The target value of low-carbon supplemental lighting based on the light requirement of plants and the corresponding decision model were determined.

[0156] (6) In order to overcome the problem of complex calculation of the correspondence between PWM signal and lamp light intensity, PID-PWM control is used to realize the light intensity control of supplementary lighting based on target light intensity.

[0157] This embodiment provides technical support for the regulation of light environment for plant growth. The system realizes the perception of plant information and supplemental light power consumption, providing key information feedback for energy-saving supplemental lighting. The supplemental lighting decision model embedded in the cloud takes the plant's net photosynthetic rate, supplemental lighting carbon emissions, and supplemental light power consumption as targets. Under the constraints of plant light requirements and supplemental light lamp current, it determines the supplemental light intensity corresponding to the light quality under different environmental conditions and plant growth stages, which is the core of low-carbon and energy-saving supplemental lighting technology. The supplemental lighting control method is based on the perception of ambient light intensity and the supplemental lighting decision model, realizing real-time light intensity regulation, and providing a feasible solution for the specific realization of low-carbon and energy-saving supplemental lighting.

[0158] Example 3

[0159] The following describes the low-carbon energy-saving supplemental lighting device based on plant light requirements provided by the present invention. The low-carbon energy-saving supplemental lighting device based on plant light requirements described below can be referred to in correspondence with the low-carbon energy-saving supplemental lighting method based on plant light requirements described above.

[0160] Figure 7 This is a schematic diagram of a low-carbon, energy-saving supplemental lighting device based on plant light requirements, as described in an embodiment of the present invention. Figure 7 As shown, the technical solution of the present invention also provides a low-carbon, energy-saving supplemental lighting device based on the light requirements of plants, the device comprising:

[0161] The acquisition module is used to acquire environmental data of the plants;

[0162] The decision module is used to input environmental data into the supplemental lighting decision model, and the supplemental lighting decision model outputs the target values ​​of light-related quantities;

[0163] The adjustment module is used to adjust the plant's supplemental light system based on the target value, so that the real-time values ​​of light-related quantities collected in the plant's environment are close to the target value.

[0164] in,

[0165] The supplemental lighting decision model can optimize and obtain the target value by minimizing the combined weighted value of the predicted negative value of photosynthetic rate, net carbon emissions from supplemental lighting, and supplemental photoelectric power.

[0166] This embodiment aims to improve the photosynthetic rate of plants, reduce supplemental lighting power and carbon emissions. It calculates the negative value of the predicted photosynthetic rate, the net carbon emissions from supplemental lighting, and the photoelectric power of supplemental lighting under the supplemental lighting scenario, and obtains the optimal supplemental lighting target value to achieve low-carbon and energy-saving supplemental lighting for plants.

[0167] Figure 8 A schematic diagram of the physical structure of an electronic device provided by the present invention, such as... Figure 8 As shown, the electronic device may include: a processor 810, a communication interface 820, a memory 830, and a communication bus 840, wherein the processor 810, the communication interface 820, and the memory 830 communicate with each other via the communication bus 840. The processor 810 can call logical instructions in the memory 830 to execute a low-carbon, energy-saving supplemental lighting method based on the plant's light requirements, the method including:

[0168] Obtain environmental data of the plants;

[0169] The environmental data is input into the supplemental lighting decision model, and the supplemental lighting decision model outputs the target value of the illumination-related quantity.

[0170] Based on the target value, the supplemental light system of the plant is adjusted so that the real-time values ​​of light-related quantities collected in the environment where the plant is located approach the target value.

[0171] in,

[0172] The supplemental lighting decision model can optimize and obtain the target value by minimizing the combined weighted value of the predicted negative photosynthetic rate, the net carbon emissions from supplemental lighting, and the photoelectric power from supplemental lighting.

[0173] Furthermore, the logical instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0174] On the other hand, the present invention also provides a computer program product, the computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions, wherein when the program instructions are executed by a computer, the computer is able to execute the low-carbon energy-saving supplemental lighting method based on the light requirements of plants provided by the above methods, the method comprising:

[0175] Obtain environmental data of the plants;

[0176] The environmental data is input into the supplemental lighting decision model, and the supplemental lighting decision model outputs the target value of the illumination-related quantity.

[0177] Based on the target value, the supplemental light system of the plant is adjusted so that the real-time values ​​of light-related quantities collected in the environment where the plant is located approach the target value.

[0178] in,

[0179] The supplemental lighting decision model can optimize and obtain the target value by minimizing the combined weighted value of the predicted negative photosynthetic rate, the net carbon emissions from supplemental lighting, and the photoelectric power from supplemental lighting.

[0180] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned low-carbon energy-saving supplemental lighting methods based on plant light requirements, the methods comprising:

[0181] Obtain environmental data of the plants;

[0182] The environmental data is input into the supplemental lighting decision model, and the supplemental lighting decision model outputs the target value of the illumination-related quantity.

[0183] Based on the target value, the supplemental light system of the plant is adjusted so that the real-time values ​​of light-related quantities collected in the environment where the plant is located approach the target value.

[0184] in,

[0185] The supplemental lighting decision model can optimize and obtain the target value by minimizing the combined weighted value of the predicted negative photosynthetic rate, the net carbon emissions from supplemental lighting, and the photoelectric power from supplemental lighting.

[0186] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0187] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0188] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A low-carbon energy-saving light supplementing method based on the light demand of a plant, characterized in that, The method comprises: obtaining environmental data of the plant; inputting the environmental data into a light supplement decision model, the light supplement decision model outputting a target value of a light-related quantity; based on the target value, adjusting a light supplement subsystem of the plant, so that a real-time value of the light-related quantity collected in the environment where the plant is located tends to approach the target value; wherein, the light supplement decision model can obtain the target value by optimizing a comprehensive weighted value of the minimum of the negative value of the predicted photosynthetic rate, the light supplement net carbon emission amount and the light supplement electric power.

2. The method for low-carbon and energy-saving light supplement based on the light demand of plants according to claim 1, characterized in that, The method for obtaining the predicted photosynthetic rate further comprises: obtaining the target value; inputting the target value into an LS-SVR model, the LS-SVR model outputting a predicted photosynthetic rate corresponding to the target value; wherein, the kernel function of the LS-SVR model adopts a radial basis function, and the formula of the radial basis function is as follows: where x i and x j correspond to two samples of the LS-SVR model, respectively, ||x i -x j || is the Euclidean distance of x i and x j , and σ is the spread constant of the radial basis function.

3. The method for low-carbon and energy-saving light supplement based on the light demand of plants according to claim 2, characterized in that, the hyperparameters of the LS-SVR model include a regularization coefficient γ and an expansion constant σ of the radial basis function; the training method of the LS-SVR model further comprises: obtaining samples for training; based on the samples, forming a test set, a validation set and a training set through cross-validation; encoding the individual position in evolutionary calculation as the hyperparameters to be optimized; obtaining the average value of the prediction error of the photosynthetic rate of the hyperparameters represented by the individual position on the validation set, taking the average value as the fitness value corresponding to the individual; updating the individual position and the fitness value corresponding to the individual until the maximum number of iterations is reached or the error is less than the set value, and obtaining the optimal hyperparameters; based on the optimal hyperparameters, completing the training of the LS-SVR model through the training set.

4. The method for low-carbon and energy-saving light supplement based on the light requirement of plants according to claim 3, characterized in that, Obtaining samples for training comprises: obtaining the light quantum flux density in the environment where the plant is located and the proportion of light waves of several different wavelengths through a light quantum sensor as the input quantity of the LS-SVR model; obtaining the actual photosynthetic rate of the plant corresponding to the light quantum flux density and the proportion of light waves through a photosynthetic rate instrument as the output quantity of the LS-SVR model; taking the input quantity and the corresponding output quantity as one sample of the LS-SVR model.

5. The method for low-carbon and energy-saving light supplement based on the light demand of plants according to claim 2, characterized in that, The method for obtaining the target value further comprises at least one of the first constraint and the second constraint; the first constraint comprises: the current of the light supplement lamp in the light supplement subsystem is less than or equal to the maximum rated current; the second constraint comprises: the target value is not less than the corresponding value of the light-related quantity of the plant at the light compensation point; the method for obtaining the corresponding value of the light-related quantity of the plant at the light compensation point further comprises: obtaining the corresponding value of the light-related quantity of the light compensation point through the bisection method based on the trained LS-SVR model.

6. The method for low-carbon and energy-saving light supplement based on the light demand of plants according to claim 1, characterized in that, The method for obtaining the light supplement net carbon emission amount further comprises: obtaining the corresponding carbon emission amount of the energy consumed by the operation of the light supplement subsystem; obtaining the amount of carbon fixed by the plant due to light supplement; subtracting the amount of carbon fixed by light supplement from the corresponding carbon emission amount of the consumed energy as the light supplement net carbon emission amount.

7. The method for low-carbon and energy-saving light supplement based on the light demand of plants according to claim 1, characterized in that, The light supplement lamp in the light supplement subsystem is controlled through a PWM signal. Based on the target value, adjust the light supplement subsystem of the plant, so that the real-time value of the light-related quantity collected in the environment where the plant is located approaches the target value, comprising: The incremental proportional integral differential algorithm is adopted to regulate the PWM signal.

8. A low-carbon energy-saving light supplementing device based on the light requirement of a plant, characterized in that, The device comprises: An acquisition module is configured to acquire environment data of a plant; A decision module is configured to input the environment data into a light supplement decision model, and the light supplement decision model outputs a target value of a light-related quantity; An adjustment module is configured to adjust a light supplement subsystem of the plant based on the target value, so that a real-time value of the light-related quantity collected in an environment where the plant is located approaches the target value; Wherein, The light supplement decision model can optimize the target value in the direction of minimizing the comprehensive weighted value of the negative value of the predicted photosynthetic rate, the light supplement net carbon emission quantity, and the light supplement electric power.

9. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the program to implement the steps of the low-carbon and energy-saving light supplement method based on the light illumination demand of the plant according to any one of claims 1-7.

10. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the low-carbon and energy-saving light supplement method based on the light illumination demand of the plant according to any one of claims 1-7.

Images