An intelligent adjustment method and system for the growth environment of greenhouse crops

Abstract

The present invention provides an intelligent adjustment method and system for the growth environment of greenhouse crops, which relates to the technical field of intelligent adjustment and includes: obtaining the current temperature parameters, humidity parameters and light parameters of greenhouse crops as the current environmental parameter set, obtaining the adjustment feasible region according to the environmental parameter set within a preset historical time range, constructing a fitness function according to cost and growth conditions, setting constraint conditions for environmental parameter optimization to obtain the optimal environmental parameter set, wherein, based on image processing recognition, obtaining the growth conditions of greenhouse crops, and using the optimal environmental parameter set to adjust the environmental parameters in the growth environment. The present invention solves the technical problems existing in the traditional method for adjusting the growth environment of greenhouse crops, such as relying on manual adjustment, being easily affected by subjective factors, resulting in low efficiency and high cost, and at the same time being unable to meet the development needs of modern agriculture for high efficiency and intelligence.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent regulation, and particularly relates to an intelligent regulation method and system for the growth environment of greenhouse crops. Background Art

[0002] With the progress of technology, greenhouse cultivation, as an efficient agricultural cultivation method, has been paid more and more attention. During the process of greenhouse cultivation, the control of environmental parameters is one of the key factors affecting crop growth and yield. Traditional control methods are usually manual adjustment, which is inefficient, costly, and easily affected by subjective factors, and cannot meet the development needs of modern agriculture for high efficiency and intelligence. Therefore, it is particularly important to study a method for intelligently regulating the growth environment of greenhouse crops. Summary of the Invention

[0003] The embodiments of the present application provide an intelligent regulation method and system for the growth environment of greenhouse crops, which are used to solve the technical problems that the traditional method for regulating the growth environment of greenhouse crops relies on manual adjustment, is easily affected by subjective factors, resulting in low efficiency and high cost, and at the same time cannot meet the development needs of modern agriculture for high efficiency and intelligence.

[0004] In view of the above problems, the embodiments of the present application provide an intelligent regulation method and system for the growth environment of greenhouse crops.

[0005] In a first aspect, the embodiments of the present application provide an intelligent regulation method for the growth environment of greenhouse crops, the method comprising: obtaining the current temperature parameter, humidity parameter, and light parameter in the growth environment of the greenhouse crops as the current environmental parameter set; obtaining the adjustment feasible region for current environmental parameter adjustment according to the environmental parameter set in the growth environment within a preset historical time range; constructing a fitness function according to the cost of adjusting environmental parameters in the growth environment and the growth situation of the greenhouse crops after adjusting the environmental parameters; setting constraint conditions according to the current environmental parameter set, and performing environmental parameter optimization within the adjustment feasible region based on the fitness function and the constraint conditions to obtain an optimal environmental parameter set, wherein the growth situation of the greenhouse crops is obtained based on image processing and recognition; and adjusting the environmental parameters in the growth environment by using the optimal environmental parameter set.

[0006] Second aspect, an embodiment of the present application provides an intelligent regulation system for the growth environment of greenhouse crops. The system includes: an environmental parameter set acquisition module for acquiring the current temperature parameter, humidity parameter, and light parameter in the growth environment of greenhouse crops as the current environmental parameter set; a regulation feasible region acquisition module for acquiring the regulation feasible region for current environmental parameter regulation according to the environmental parameter sets in the growth environment within a preset historical time range; a fitness function construction module for constructing a fitness function according to the cost of adjusting environmental parameters in the growth environment and the growth situation of the greenhouse crops after adjusting the environmental parameters; an environmental parameter optimization module for setting constraint conditions according to the current environmental parameter set, and performing environmental parameter optimization within the regulation feasible region based on the fitness function and the constraint conditions to obtain an optimal environmental parameter set, wherein the growth situation of the greenhouse crops is acquired based on image processing and recognition; and an environmental parameter regulation module for regulating the environmental parameters in the growth environment by using the optimal environmental parameter set.

[0007] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0008] Acquire the current temperature parameter, humidity parameter, and light parameter in the growth environment of greenhouse crops as the current environmental parameter set, acquire the regulation feasible region according to the environmental parameter sets in the growth environment within a preset historical time range, construct a fitness function according to the cost of adjusting environmental parameters in the growth environment and the growth situation of the greenhouse crops after adjusting the environmental parameters, set constraint conditions, perform environmental parameter optimization to obtain an optimal environmental parameter set, wherein the growth situation of the greenhouse crops is acquired based on image processing and recognition, and regulate the environmental parameters in the growth environment by using the optimal environmental parameter set.

[0009] It solves the technical problem that the traditional method for regulating the growth environment of greenhouse crops relies on manual regulation, is easily affected by subjective factors, resulting in low efficiency and high cost, and at the same time cannot meet the development needs of modern agriculture for high efficiency and intelligence. It realizes automatically acquiring environmental parameters, constructing a regulation feasible region and a fitness function according to historical data, thereby realizing the optimization of environmental parameters, achieving the technical effect of effectively regulating the growth environment of greenhouse crops, reducing production costs, and at the same time realizing intelligent production.

[0010] The above description is only an overview of the technical solution of the present application. In order to be able to understand the technical means of the present application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features, and advantages of the present application more obvious and understandable, the following specifically gives the specific implementation manners of the present application. Brief Description of the Drawings

[0011] To more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings of the embodiments of the present disclosure will be briefly introduced below. Obviously, the drawings in the following description only relate to some embodiments of the present disclosure and do not limit the present disclosure.

[0012] Figure 1 FIG. is a schematic flow chart of an intelligent adjustment method for the growth environment of greenhouse crops provided by an embodiment of the present application;

[0013] Figure 2 FIG. is a schematic flow chart of obtaining an adjustment feasible region in an intelligent adjustment method for the growth environment of greenhouse crops provided by an embodiment of the present application;

[0014] Figure 3 FIG. is a schematic flow chart of obtaining an optimal environmental parameter set in an intelligent adjustment method for the growth environment of greenhouse crops provided by an embodiment of the present application;

[0015] Figure 4 FIG. is a schematic structural diagram of an intelligent adjustment system for the growth environment of greenhouse crops provided by an embodiment of the present application.

[0016] Description of the reference numerals: The environmental parameter set acquisition module 10, the adjustment feasible region acquisition module 20, the fitness function construction module 30, the environmental parameter optimization module 40, and the environmental parameter adjustment module 50. Detailed Embodiments

[0017] By providing an intelligent adjustment method for the growth environment of greenhouse crops, the embodiments of the present application solve the technical problems existing in the traditional adjustment method for the growth environment of greenhouse crops, which rely on manual adjustment and are easily affected by subjective factors, resulting in low efficiency and high cost, and at the same time cannot meet the development needs of modern agriculture for high efficiency and intelligence. The technical effects of automatically obtaining environmental parameters, constructing an adjustment feasible region and a fitness function based on historical data, thereby realizing the optimization of environmental parameters, effectively adjusting the growth environment of greenhouse crops, reducing production costs, and realizing intelligent production are achieved.

[0018] Embodiment 1

[0019] As Figure 1 shown, an embodiment of the present application provides an intelligent adjustment method for the growth environment of greenhouse crops, and the method includes:

[0020] Step S100: Obtain the current temperature parameter, humidity parameter, and light parameter in the growth environment of the greenhouse crops as the current environmental parameter set;

[0021] Specifically, temperature, humidity, and light sensors are deployed inside the greenhouse. The sensors send data to the central processing system through wireless connections, monitor and collect data on these parameters in real time, integrate the obtained temperature, humidity, and light data together to form the current set of environmental parameters, which serves as the basis for subsequent step analysis and adjustment. This realizes clarifying the current environmental conditions inside the greenhouse so as to make corresponding adjustments according to the crop requirements.

[0022] Step S200: Obtain the adjustable feasible region for the current environmental parameter adjustment according to the set of environmental parameters in the growth environment within a preset historical time range.

[0023] Furthermore, as Figure 2 shown, step S200 of this application further includes:

[0024] Step S210: Obtain multiple sets of historical environmental parameters in the growth environment within a preset historical time range, and obtain a set of historical temperature parameters, a set of historical humidity parameters, and a set of historical light parameters.

[0025] Step S220: Perform error compensation on the historical temperature parameters, historical humidity parameters, and historical light parameters in the set of historical temperature parameters, the set of historical humidity parameters, and the set of historical light parameters according to a preset ratio to obtain multiple ranges of historical temperature parameters, multiple ranges of historical light parameters, and multiple ranges of historical humidity parameters.

[0026] Step S230: Randomly generate temperature parameters, humidity parameters, and light parameters in the multiple ranges of historical temperature parameters, multiple ranges of historical light parameters, and multiple ranges of historical humidity parameters respectively, combine them to obtain multiple sets of environmental parameters, and obtain the adjustable feasible region.

[0027] Specifically, analyze the historical environmental parameter data to determine the feasible range of the current environmental parameter adjustment. According to the growth cycle of the crop and the availability of the data, preset the historical time range, such as the past few days, weeks, or months. Extract multiple sets of historical environmental parameters within the preset historical time range from the database or storage system, including the records of temperature, humidity, and light parameters over a period of time in the past. Classify the extracted historical environmental parameter data according to temperature, humidity, and light, and integrate them into a set of historical temperature parameters, a set of historical humidity parameters, and a set of historical light parameters respectively. These sets are used to clarify the changes of each environmental parameter within the historical time range.

[0028] According to the growth requirements of crops and practical experience, a suitable preset ratio is set, such as ±10%. This ratio is used for error compensation of historical environmental parameters. For the historical temperature parameter set, historical humidity parameter set, and historical light parameter set, the error compensation range of each historical parameter is calculated according to the preset ratio. For example, if the historical temperature parameter is 20 degrees, and the compensation is carried out according to the preset ratio of ±10%, the adjusted range is [18, 22] degrees. According to the error compensation results, multiple historical temperature parameter ranges, multiple historical light parameter ranges, and multiple historical humidity parameter ranges are obtained respectively. These parameter ranges are used to construct the adjustment feasible region, providing a reference for environmental parameter adjustment.

[0029] Within each historical temperature parameter range, historical light parameter range, and historical humidity parameter range, temperature parameters, light parameters, and humidity parameters are randomly generated respectively, and they are randomly combined to form multiple environmental parameter sets. These sets represent various environmental conditions that may occur within the adjustment feasible region. The adjustment feasible region composed of multiple environmental parameter sets provides a search space for subsequent environmental parameter optimization. Within this feasible region, the system will perform optimization according to the fitness function and constraint conditions to find the best environmental parameter set.

[0030] Step S300: Construct a fitness function according to the cost of adjusting environmental parameters in the growth environment and the growth situation of greenhouse crops after adjusting environmental parameters;

[0031] Specifically, cost factors are determined, including energy consumption, water resource consumption, and labor cost, etc. Among them, energy consumption includes heating, cooling, lighting, etc. These factors are quantified into a cost index to measure the economic effect of environmental parameter adjustment;

[0032] Photos of greenhouse crops are taken using a camera or other image acquisition devices, and these images are processed and analyzed to identify the growth situation of the crops, such as leaf area, stem diameter, number of inflorescences, etc. This information is used to more accurately evaluate the growth effect of environmental parameter adjustment. Growth effect factors are determined, including crop growth rate, growth cycle, and yield index, etc. These factors are quantified into a growth effect index to measure the impact of environmental parameter adjustment on crop growth.

[0033] The cost index and growth effect index are integrated into a fitness function. The fitness function is used to minimize the cost index while maximizing the growth effect index. The fitness function can be adjusted and optimized according to the actual situation to meet the needs of different crops and greenhouse conditions.

[0034] The fitness function is as follows:

[0035]

[0036] Among them, Y is the fitness, and W K , W S and W C are weights, K A is the cost of the adjusted set of environmental parameters, and K B is the cost of the current set of environmental parameters. is the size score of the j-th greenhouse crop in the growth environment after adjusting the set of environmental parameters. T is the number of detected greenhouse crops, and S B is the average size score of the greenhouse crops in the growth environment under the current set of environmental parameters. is the color score of the j-th greenhouse crop in the growth environment after adjusting the set of environmental parameters, and C B is the average color score of the greenhouse crops in the growth environment under the current set of environmental parameters. The color score and the size score are obtained through image processing and recognition.

[0037] Step S400: According to the current set of environmental parameters, set constraint conditions. Based on the fitness function and the constraint conditions, perform optimization of environmental parameters within the adjustment feasible region to obtain the optimal set of environmental parameters. Among them, based on image processing and recognition, obtain the growth situation of the greenhouse crops.

[0038] Furthermore, as Figure 3 shown, step S400 of this application further includes:

[0039] Step S410: Set the fitness greater than 0 as a constraint condition. During the optimization process, if the set of environmental parameters does not meet the constraint condition, it is discarded.

[0040] Step S420: Randomly select a set of environmental parameters within the adjustment feasible region as the first set of environmental parameters and temporarily use it as the optimal solution.

[0041] Step S430: According to the fitness function, obtain the first fitness of the first set of environmental parameters.

[0042] Step S440: Randomly select another set of environmental parameters within the adjustment feasible region as the second set of environmental parameters and obtain the second fitness according to the fitness function.

[0043] Step S450: Determine whether the second fitness is greater than the first fitness. If so, replace the second set of environmental parameters as the optimal solution. If not, randomly generate a random number within [0, 1] and determine whether this random number is less than a probability value. If so, use the second set of environmental parameters as the optimal solution. If not, still use the first set of environmental parameters as the optimal solution. The probability value decreases as the number of iterations increases.

[0044] Step S460: Continue the iterative optimization until the first optimization iteration count is reached. When the fitness of the newly obtained set of environmental parameters in the iteration is less than or equal to the fitness of the optimal solution, add the new set of environmental parameters to the taboo list.

[0045] Step S470: Continue the iterative optimization until the second optimization iteration count is reached, output the final optimal solution, and obtain the optimal set of environmental parameters.

[0046] Specifically, according to the current set of environmental parameters, set the constraint conditions. To ensure that the set of environmental parameters has a positive impact on the growth of greenhouse crops, set the fitness greater than 0 as the constraint condition. During the optimization process, calculate the fitness value of each set of environmental parameters according to the fitness function. If a set of environmental parameters with a fitness not greater than 0 is encountered, that is, it does not meet the constraint conditions, discard it and do not consider it.

[0047] Within the adjustable feasible region, randomly select a set of environmental parameters as the initial solution. This initial solution includes temperature parameters, humidity parameters, and light parameters, and use this as the starting point of the optimization process. Temporarily use the selected initial solution as the optimal solution for comparison and update in the subsequent optimization process.

[0048] Substitute the first set of environmental parameters into the fitness function to calculate its fitness value, that is, the first fitness. This first fitness reflects the degree of adaptation of the initial solution to the growth of greenhouse crops and serves as a reference for subsequent optimization. Within the adjustable feasible region, randomly select another set of environmental parameters as the second set of environmental parameters for comparison with the first set of environmental parameters. Substitute the second set of environmental parameters into the fitness function to calculate its fitness value, that is, the second fitness. This second fitness reflects the degree of adaptation of the second set of environmental parameters to the growth of greenhouse crops.

[0049] Compare the first fitness and the second fitness. If the second fitness is greater than the first fitness, then replace the second set of environmental parameters corresponding to the second fitness as the optimal solution. If the second fitness is not greater than the first fitness, generate a random number within the interval [0, 1] and compare it with the preset probability value. The probability value decreases as the number of iterations increases to reduce the possibility of accepting suboptimal solutions in the later iterations. Among them, the probability value is calculated by the following formula:

[0050]

[0051] Where Y 2 is the second fitness, Y 1 is the first fitness, and N is a constant that decreases as the number of iterations increases.

[0052] If the random number is less than the probability value, the second set of environmental parameters is taken as the optimal solution; otherwise, the first set of environmental parameters is still taken as the optimal solution.

[0053] Reaching the first optimization iteration count is equivalent to the optimization in the first stage. In this stage, continuous iteration is carried out to find a better set of environmental parameters. For solutions with lower fitness, it is judged whether to accept them as the optimal solution according to the probability value, so as to improve the optimization efficiency and jump out of the local optimum. During the iteration process, when the fitness of the newly obtained set of environmental parameters is less than or equal to the fitness of the current optimal solution, this newly obtained set of environmental parameters is added to the taboo list. The taboo list is a strategy to avoid repeated search and falling into the local optimal solution. It records the non-optimal solutions that have been searched to avoid repeated search for these solutions in subsequent iterations.

[0054] Reaching the second optimization iteration count is equivalent to the optimization in the second stage. In this stage, continuous iteration is continued, and the probability value and the taboo list are adjusted. For solutions with lower fitness, they are directly discarded and tabued to improve the optimization accuracy, and the found optimal solution is output as the optimal set of environmental parameters.

[0055] Furthermore, step S430 of the present application further includes:

[0056] Step S431: Obtain T crop images of T greenhouse crops in the growth environment under the first set of environmental parameters;

[0057] Step S432: Based on the greenhouse crop growth monitoring data within the historical time in multiple growth environments, construct a growth situation analysis model, where the growth situation analysis model includes a size analysis unit and a color analysis unit;

[0058] Step S433: Input the T crop images into the growth situation analysis model respectively to obtain T size scores and T color scores;

[0059] Step S434: Obtain the greenhouse environment cost under the first set of environmental parameters;

[0060] Step S435: Input the T size scores, the T color scores, and the greenhouse environment cost into the fitness function to obtain the first fitness.

[0061] Specifically, under the first set of environmental parameters, T greenhouse crops in the growth environment are photographed by a camera or other image acquisition devices installed in the greenhouse to obtain their crop images, where T is the number of greenhouse crops. For example, there are 30 greenhouse crops in the growth environment, and image acquisition is carried out for them respectively.

[0062] Collect the greenhouse crop growth monitoring data of historical time in multiple growth environments. These data include characteristics such as the size, color, and growth rate of greenhouse crops. Using the collected data, build a growth situation analysis model based on a convolutional neural network. The model includes a size analysis unit and a color analysis unit. The size analysis unit is used to analyze the size characteristics of greenhouse crops, such as plant height, leaf length and width, etc.; the color analysis unit is used to analyze the color characteristics of greenhouse crops, such as leaf color, flower color, etc. Use the historical data to train the growth situation analysis model and verify it to ensure that the model can accurately analyze the growth situation of greenhouse crops.

[0063] Input the obtained T crop images into the growth situation analysis model respectively. In the model, the size analysis unit conducts size feature analysis on the input T crop images, and calculates a size score for each crop image based on size characteristics such as plant height, leaf length and width; the color analysis unit conducts color feature analysis on the input T crop images, and calculates a color score for each crop image based on color characteristics such as leaf color, flower color.

[0064] The factors affecting the greenhouse environment cost include energy consumption (such as electricity, water resources, etc.), equipment maintenance costs, labor costs, etc. Calculate the cost according to factors such as energy consumption and equipment maintenance costs under the first set of environmental parameters, and obtain the greenhouse environment cost under this set of environmental parameters. Substitute the T size scores, the T color scores and the greenhouse environment cost into the fitness function to calculate the fitness value corresponding to the first set of environmental parameters, that is, the first fitness.

[0065] Furthermore, step S432 of this application further includes:

[0066] Step S4321: Based on the greenhouse crop growth monitoring data of historical time in multiple growth environments, obtain a set of historical crop images, a set of historical size scores, and a set of historical color scores;

[0067] Step S4322: Based on a convolutional neural network, build the size analysis unit and the color analysis unit;

[0068] Step S4323: Use the set of historical crop images and the set of historical size scores to conduct supervised training, verification, and testing on the size analysis unit, and update the network parameters until the first convergence condition is reached to obtain the size analysis unit;

[0069] Step S4324: Use the set of historical crop images and the set of historical color scores to conduct supervised training, verification, and testing on the color analysis unit, and update the network parameters until the second convergence condition is reached to obtain the color analysis unit and obtain the growth situation analysis model.

[0070] Specifically, growth monitoring data of greenhouse crops in different growth environments are collected, including crop images, size scores, and color scores. Among them, the size scores and color scores are obtained by those skilled in the art through evaluating the crop images. For example, in the case of cucumbers, the size score can be based on the length and thickness of the cucumbers, and the color score can be evaluated based on the color and surface texture of the cucumbers. The collected crop images, size scores, and color scores are classified and sorted to form a historical crop image set, a historical size score set, and a historical color score set.

[0071] A convolutional neural network (CNN) is used to construct a size analysis unit and a color analysis unit. A convolutional neural network is a deep learning technology suitable for image processing and recognition tasks. Based on the convolutional neural network, the network structure of the analysis unit is constructed, including convolutional layers, activation functions, pooling layers, fully connected layers, etc. Among them, the size analysis unit and the color analysis unit can share some convolutional layers and output size scores and color scores respectively in the subsequent fully connected layers.

[0072] The historical crop image set and the historical size score set are randomly divided into a training set, a validation set, and a test set according to a certain ratio. For example, they are divided into a 70% training set, a 15% validation set, and a 15% test set. The crop images in the training set are used as inputs, and the corresponding historical size scores are used as output targets. The size analysis unit is trained using a supervised learning method. During the training process, the network parameters are continuously updated to minimize the prediction error. After the training is completed, the performance of the size analysis unit is evaluated using the validation set, and hyperparameters such as the network structure and learning rate are adjusted according to the validation results to improve the model performance.

[0073] A first convergence condition is set, such as the training loss being less than a certain threshold, the validation loss no longer decreasing significantly, or reaching the maximum training cycle, etc. When the convergence condition is met, the training process is stopped. The final performance of the size analysis unit is evaluated using the test set, and the test results are used as an estimate of the model's performance on unknown data.

[0074] In the same way, the color analysis unit is constructed using the historical crop image set and the historical color score set. For the sake of brevity of the specification, it will not be elaborated here. The size analysis unit and the color analysis unit are integrated to obtain the growth situation analysis model.

[0075] Step S500: Adjust the environmental parameters in the growth environment by using the optimal environmental parameter set.

[0076] Specifically, control devices related to temperature, humidity, and lighting adjustment, such as heaters, refrigeration equipment, humidity controllers, lighting systems, etc., are connected to the central control system. According to the optimal environmental parameter set, corresponding control instructions are generated for each control device. The control device adjusts its output according to the received instructions to change the temperature, humidity, and lighting conditions in the greenhouse, and adjusts them to the parameters within the optimal environmental parameter set, so as to create a more favorable environment for crop growth and development while controlling costs.

[0077] In summary, the intelligent adjustment method and system for the growth environment of greenhouse crops provided by the embodiments of the present application have the following technical effects:

[0078] Obtain the current temperature parameter, humidity parameter, and lighting parameter in the growth environment of greenhouse crops as the current environmental parameter set. According to the environmental parameter set in the growth environment within a preset historical time range, obtain the adjustment feasible region. According to the cost of adjusting environmental parameters in the growth environment and the growth situation of greenhouse crops after adjusting environmental parameters, construct a fitness function, set constraint conditions, perform environmental parameter optimization, and obtain the optimal environmental parameter set. Among them, based on image processing and recognition, obtain the growth situation of greenhouse crops, and use the optimal environmental parameter set to adjust the environmental parameters in the growth environment.

[0079] It solves the technical problems that the traditional method for adjusting the growth environment of greenhouse crops relies on manual adjustment, is easily affected by subjective factors, resulting in low efficiency and high costs, and at the same time cannot meet the development needs of modern agriculture for high efficiency and intelligence. It realizes the automatic acquisition of environmental parameters, constructs an adjustment feasible region and a fitness function based on historical data, thereby realizing the optimization of environmental parameters, achieving the technical effect of effectively adjusting the growth environment of greenhouse crops, reducing production costs, and realizing intelligent production at the same time.

[0080] Embodiment 2

[0081] Based on the same inventive concept as the intelligent adjustment method for the growth environment of greenhouse crops in the foregoing embodiment, as Figure 4 shown, the present application provides an intelligent adjustment system for the growth environment of greenhouse crops, and the system includes:

[0082] An environmental parameter set acquisition module 10, which is used to acquire the current temperature parameter, humidity parameter, and lighting parameter in the growth environment of greenhouse crops as the current environmental parameter set;

[0083] An adjustment feasible region acquisition module 20, which is used to acquire the adjustment feasible region for current environmental parameter adjustment according to the environmental parameter set in the growth environment within a preset historical time range;

[0084] A fitness function construction module 30, which is used to construct a fitness function according to the cost of adjusting environmental parameters in the growth environment and the growth conditions of the greenhouse crops after adjusting the environmental parameters;

[0085] An environmental parameter optimization module 40, which is used to set constraint conditions according to the current environmental parameter set, and perform environmental parameter optimization within the adjustment feasible region based on the fitness function and the constraint conditions to obtain an optimal environmental parameter set, wherein the growth conditions of the greenhouse crops are obtained based on image processing and recognition;

[0086] An environmental parameter adjustment module 50, which is used to adjust the environmental parameters in the growth environment by using the optimal environmental parameter set.

[0087] Furthermore, the system further includes:

[0088] A historical environmental parameter acquisition module, which is used to acquire multiple historical environmental parameter sets within a preset historical time range of the growth environment, and obtain a historical temperature parameter set, a historical humidity parameter set, and a historical light parameter set;

[0089] An error compensation module, which is used to perform error compensation on the historical temperature parameters, historical humidity parameters, and historical light parameters in the historical temperature parameter set, historical humidity parameter set, and historical light parameter set according to a preset ratio to obtain multiple historical temperature parameter ranges, multiple historical light parameter ranges, and multiple historical humidity parameter ranges;

[0090] An environmental parameter acquisition module, which is used to randomly generate temperature parameters, humidity parameters, and light parameters within the multiple historical temperature parameter ranges, multiple historical light parameter ranges, and multiple historical humidity parameter ranges respectively, combine them to obtain multiple environmental parameter sets, and obtain the adjustment feasible region.

[0091] Furthermore, the system further includes:

[0092] Construct a fitness function as follows:

[0093]

[0094] Where Y is the fitness, W K 、W S and W C are weights, K A is the cost of the adjusted environmental parameter set, K B is the cost of the current environmental parameter set, is the size score of the jth greenhouse crop in the growth environment after adjusting the environmental parameter set, T is the number of detected greenhouse crops, SB the average size score of the greenhouse crops in the growth environment under the current environmental parameter set is the color score of the j-th greenhouse crop in the growth environment after adjusting the environmental parameter set, C B is the average color score of the greenhouse crops in the growth environment under the current environmental parameter set. The color score and the size score are obtained through image processing and recognition.

[0095] Furthermore, the system further includes:

[0096] a rejection module, which is used to set the fitness greater than 0 as a constraint condition. During the optimization process, if the environmental parameter set does not meet the constraint condition, it will be rejected;

[0097] a temporary optimal solution acquisition module, which is used to randomly select an environmental parameter set within the adjustment feasible region as the first environmental parameter set and temporarily use it as the optimal solution;

[0098] a first fitness acquisition module, which is used to obtain the first fitness of the first environmental parameter set according to the fitness function;

[0099] a second fitness acquisition module, which is used to randomly select another environmental parameter set within the adjustment feasible region as the second environmental parameter set and obtain the second fitness according to the fitness function;

[0100] a second fitness judgment module, which is used to judge whether the second fitness is greater than the first fitness. If so, the second environmental parameter set is used as the optimal solution instead. If not, a random number within [0, 1] is randomly generated to judge whether the random number is less than a probability value. If so, the second environmental parameter set is used as the optimal solution. If not, the first environmental parameter set is still used as the optimal solution. The probability value decreases as the number of iterations increases;

[0101] a first iterative optimization module, which is used to continue iterative optimization until the first optimization iteration number is reached. When the fitness of the newly obtained environmental parameter set in the iteration is less than or equal to the fitness of the optimal solution, the new environmental parameter set is added to the taboo list;

[0102] a second iterative optimization module, which is used to continue iterative optimization until the second optimization iteration number is reached, output the final optimal solution, and obtain the optimal environmental parameter set.

[0103] Furthermore, the system further includes:

[0104] a crop image acquisition module, which is used to acquire T crop images of T greenhouse crops in the growth environment under the first environmental parameter set;

[0105] An analysis model construction module, configured to construct a growth condition analysis model based on greenhouse crop growth monitoring data within a historical time in multiple growth environments, wherein the growth condition analysis model includes a size analysis unit and a color analysis unit;

[0106] A score acquisition module, configured to input the T crop images into the growth condition analysis model respectively to obtain T size scores and T color scores;

[0107] A greenhouse environment cost acquisition module, configured to acquire the greenhouse environment cost under the first environmental parameter set;

[0108] A first fitness output module, configured to input the T size scores, the T color scores and the greenhouse environment cost into the fitness function to obtain the first fitness.

[0109] Furthermore, the system further includes:

[0110] A historical information acquisition module, configured to acquire a historical crop image set, a historical size score set and a historical color score set based on greenhouse crop growth monitoring data within a historical time in multiple growth environments;

[0111] An analysis unit construction module, configured to construct the size analysis unit and the color analysis unit based on a convolutional neural network;

[0112] A size analysis unit training module, configured to perform supervised training, verification and testing on the size analysis unit by using the historical crop image set and the historical size score set, and update network parameters until a first convergence condition is reached to obtain the size analysis unit;

[0113] A color analysis unit training module, configured to perform supervised training, verification and testing on the color analysis unit by using the historical crop image set and the historical color score set, and update network parameters until a second convergence condition is reached to obtain the color analysis unit and obtain the growth condition analysis model.

[0114] Furthermore, the system further includes:

[0115] The probability value is calculated by the following formula:

[0116]

[0117] where Y 2 is the second fitness, Y 1 is the first fitness, and N is a constant that decreases as the number of iterations increases.

[0118] Through the foregoing detailed description of a method for intelligently adjusting the growth environment of greenhouse crops, those skilled in the art can clearly understand a method and system for intelligently adjusting the growth environment of greenhouse crops in this embodiment. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple. For related parts, reference can be made to the description in the method part.

[0119] The foregoing description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent adjustment method for the growth environment of greenhouse crops, characterized in that, the method includes: Obtaining the current temperature parameter, humidity parameter and light parameter in the growth environment of greenhouse crops as the current environmental parameter set; Obtaining the adjustable feasible region for current environmental parameter adjustment according to the environmental parameter set in the growth environment within a preset historical time range; Constructing a fitness function according to the cost of adjusting environmental parameters in the growth environment and the growth situation of the greenhouse crops after adjusting environmental parameters; Setting constraint conditions according to the current environmental parameter set, and performing environmental parameter optimization within the adjustable feasible region based on the fitness function and constraint conditions to obtain the optimal environmental parameter set, wherein, based on image processing recognition, obtaining the growth situation of the greenhouse crops; Adjusting the environmental parameters in the growth environment by using the optimal environmental parameter set; Obtaining the temperature parameter range, humidity parameter range and light parameter range in the growth environment, and combining with the environmental parameter set in the growth environment within a preset historical time range to obtain the adjustable feasible region for current environmental parameter adjustment, including: Obtaining multiple historical environmental parameter sets in the growth environment within a preset historical time range, and obtaining a historical temperature parameter set, a historical humidity parameter set and a historical light parameter set; Performing error compensation on the historical temperature parameters, historical humidity parameters and historical light parameters in the historical temperature parameter set, historical humidity parameter set and historical light parameter set according to a preset ratio to obtain multiple historical temperature parameter ranges, multiple historical light parameter ranges and multiple historical humidity parameter ranges; Randomly generating temperature parameters, humidity parameters and light parameters in the multiple historical temperature parameter ranges, multiple historical light parameter ranges and multiple historical humidity parameter ranges respectively, and combining them to obtain multiple environmental parameter sets to obtain the adjustable feasible region; Constructing a fitness function according to the cost of adjusting environmental parameters in the growth environment and the growth situation of the greenhouse crops after adjustment, as follows: Among them, Y is the fitness, W K , W S and W C are weights, K A is the cost of the adjusted set of environmental parameters, K B is the cost of the current set of environmental parameters, is the size score of the j-th greenhouse crop in the growth environment after adjusting the set of environmental parameters, T is the number of greenhouse crops detected, S B is the average size score of the greenhouse crops in the growth environment under the current set of environmental parameters, is the color score of the j-th greenhouse crop in the growth environment after adjusting the set of environmental parameters, C B is the average color score of the greenhouse crops in the growth environment under the current set of environmental parameters. The color score and the size score are obtained by image processing recognition; Setting constraint conditions according to the current environmental parameter set, and performing environmental parameter optimization within the adjustable feasible region based on the fitness function and constraint conditions, including: Setting that the fitness is greater than 0 as a constraint condition, and during the optimization process, if the environmental parameter set does not meet the constraint condition, it is discarded; Randomly selecting an environmental parameter set within the adjustable feasible region as the first environmental parameter set and temporarily taking it as the optimal solution; Obtaining the first fitness of the first environmental parameter set according to the fitness function; Randomly selecting another environmental parameter set within the adjustable feasible region as the second environmental parameter set and obtaining the second fitness according to the fitness function; Determine whether the second fitness is greater than the first fitness. If so, replace the second set of environmental parameters as the optimal solution. If not, generate a random number within [0, 1] randomly, and determine whether the random number is less than a probability value. If so, use the second set of environmental parameters as the optimal solution. If not, still use the first set of environmental parameters as the optimal solution, and the probability value decreases as the number of iterations increases; Continue iterative optimization until the first optimization iteration number is reached. When the fitness of the newly obtained set of environmental parameters in the iteration is less than or equal to the fitness of the optimal solution, add the new set of environmental parameters to the taboo list; Continue iterative optimization until the second optimization iteration number is reached, output the final optimal solution, and obtain the optimal set of environmental parameters.

2. The method according to claim 1, characterized in that, According to the fitness function, obtaining the first fitness of the first set of environmental parameters includes: Obtain T crop images of T greenhouse crops in the growth environment under the first set of environmental parameters; Based on the greenhouse crop growth monitoring data within a historical time in multiple growth environments, construct a growth situation analysis model, wherein the growth situation analysis model includes a size analysis unit and a color analysis unit; Input the T crop images into the growth situation analysis model respectively to obtain T size scores and T color scores; Obtain the greenhouse environment cost under the first set of environmental parameters; Input the T size scores, the T color scores, and the greenhouse environment cost into the fitness function to obtain the first fitness.

3. The method according to claim 2, characterized in that, Based on the greenhouse crop growth monitoring data within a historical time in multiple growth environments, constructing a growth situation analysis model includes: Based on the greenhouse crop growth monitoring data within a historical time in multiple growth environments, obtain a historical crop image set, a historical size score set, and a historical color score set; Based on a convolutional neural network, construct the size analysis unit and the color analysis unit; Use the historical crop image set and the historical size score set to perform supervised training, verification, and testing on the size analysis unit, and update the network parameters until the first convergence condition is reached to obtain the size analysis unit; Use the historical crop image set and the historical color score set to perform supervised training, verification, and testing on the color analysis unit, and update the network parameters until the second convergence condition is reached to obtain the color analysis unit, and obtain the growth situation analysis model.

4. The method according to claim 1, characterized in that, The probability value is calculated by the following formula: Among them, Y 2 is the second fitness, Y 1 is the first fitness, and N is a constant that decreases as the number of iterations increases.

5. A greenhouse crop growth environment intelligent regulation system, characterized in that, The system is used to execute the method described in any one of claims 1-4, and the system includes: An environmental parameter set acquisition module, which is used to acquire the current temperature parameter, humidity parameter, and light parameter in the growth environment of greenhouse crops as the current environmental parameter set; Adjustable Feasible Region Acquisition Module, which is used to obtain the adjustable feasible region for current environmental parameter adjustment according to the set of environmental parameters of the growth environment within a preset historical time range; Fitness Function Construction Module, which is used to construct a fitness function according to the cost of adjusting environmental parameters in the growth environment and the growth situation of the greenhouse crops after adjusting the environmental parameters, wherein the growth situation of the greenhouse crops is obtained based on image processing and recognition; Environmental Parameter Optimization Module, which is used to set constraint conditions according to the current set of environmental parameters, and perform environmental parameter optimization within the adjustable feasible region based on the fitness function and the constraint conditions to obtain the optimal set of environmental parameters; Environmental Parameter Adjustment Module, which is used to adjust the environmental parameters in the growth environment by using the optimal set of environmental parameters.

Images