Greenhouse CO2 concentration zoning regulation and control method based on fuzzy logic reasoning

Abstract

The invention discloses a greenhouse CO2 concentration partition regulation and control method based on fuzzy logic reasoning. The method comprises the following steps: S1, acquiring an environment parameter data set corresponding to each partition in real time through a sensor network deployed in a greenhouse; s2, performing data preprocessing on the acquired environment parameter data set of each partition; s3, determining priority weights of the partitions; s4, based on a greenhouse partition priority model and the preprocessed environment parameter data set of each partition; s5, calculating the target CO2 concentration range and the regulation and control amount of each subarea; s6, converting the target CO2 concentration range and the regulation and control quantity of each subarea into executable control signals, and executing CO2 concentration subarea regulation and control; and S7, monitoring the environmental parameter data of each partition in real time through the sensor network until the target range is stable. The adaptability to the crop growth environment is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of greenhouses, and particularly to a method for regulating the concentration zoning of a greenhouse based on fuzzy logic reasoning Background Art

[0002] With the development of modern agricultural technology, intelligent greenhouses have become an important means to improve agricultural production efficiency. Among them Scientific management of the concentration plays a crucial role in the efficient growth of crops As a key element of photosynthesis, its concentration directly affects the photosynthetic rate and yield of crops. How to efficiently and accurately regulate the concentration distribution in each area inside the greenhouse is still a technical problem in the intelligent greenhouse control system

[0003] Currently, most greenhouses adopt an integral concentration control method, and uniformly release through a single supply system to the entire greenhouse. The realization of this method usually depends on a simple threshold control or a fixed-time supply scheme. By presetting the concentration target range of the whole greenhouse, when the actual concentration is lower than the lower limit, the supply equipment is triggered to run and stops supplying when it reaches the upper limit. Although it is simple and easy to implement, it has significant technical limitations

[0004] On the one hand, due to differences in crop types, growth stages, and environmental factors in different areas of the greenhouse, the requirements often vary significantly. The existing overall regulation method is difficult to adapt to the changes in the zoning requirements, and it is easy to cause the concentration in some areas to be too high or too low, affecting the photosynthesis efficiency and growth rate of crops. On the other hand, the response of the traditional control system is relatively lagging and cannot monitor and adjust the concentration distribution in real time. In the case of dynamic environmental changes, simple threshold control is easy to cause waste of resources, and excessive supply is still carried out when the demand has been met. In addition, traditional methods are difficult to quickly respond to complex environmental changes and cannot meet the requirements of precise control for efficient agricultural production

[0005] In summary, the existing technology has significant deficiencies in the precise regulation of the greenhouse concentration zoning, mainly reflected in the difficulties in realizing differential management between regions, response lag, and low resource utilization efficiency. The defects of traditional technologies directly affect the growth efficiency of crops and the operating cost of greenhouses, and there is an urgent need for a new method based on intelligent technology to solve the above problems Summary of the Invention ​

[0006] An object of the present invention is to provide a greenhouse concentration zoning regulation method based on fuzzy logic reasoning, which improves the adaptability to the crop growth environment.

[0007] A greenhouse concentration zoning regulation method based on fuzzy logic reasoning according to an embodiment of the present invention includes the following steps: S1. Real-time collect the environmental parameter datasets corresponding to each partition through a sensor network deployed inside the greenhouse; S2. Perform data preprocessing on the collected environmental parameter datasets of each partition; S3. Construct a greenhouse partition priority model according to the types, growth stages of crops in the greenhouse and the target yield requirements. The partition priority model determines the priority weights of the partitions by evaluating the importance and response sensitivity of the crops in each partition; S4. Based on the greenhouse partition priority model and the preprocessed environmental parameter datasets of each partition, construct a concentration regulation rule base in combination with a fuzzy logic inference system, The concentration regulation rule base includes the definition of membership functions, the setting of fuzzy rules and the correlation relationship between input and output variables; S5. Use the fuzzy logic inference system to calculate the target concentration range and regulation amount of each partition according to the real-time environmental parameter datasets of each partition and the priority weights of the greenhouse partition priority model; S6. Convert the target concentration range and regulation amount of each partition into executable control signals, and transmit them to the partition execution supply equipment, air flow adjustment equipment and ventilation system in the greenhouse to perform concentration zoning regulation; S7. Real-time monitor the environmental parameter data of each partition through the sensor network, evaluate whether the actual concentration reaches the target range. If the actual concentration deviates from the target range, restart the fuzzy logic inference system for calculation based on the deviation value, and update the control signal until the target range is stable.

[0008] Optionally, the specific steps of S1 include: S11. Deploy multi-type sensor nodes in each partition through the sensor network inside the greenhouse to collect the environmental parameter data of each partition in real time. The environmental parameter data includes concentration and temperature and humidity , light intensity , air flow distribution characteristics and plant growth stage , where i represents the partition number and j represents the time series index; S12. Construct the environmental parameter data set D from the collected environmental parameter data of each partition: ; where m represents the total number of partitions in the greenhouse and n represents the total number of time series sampling points.

[0009] Optionally, the specific steps of S2 include: S21. Denoise each element in the collected environmental parameter data set D and calculate the denoised value using the moving average filtering algorithm: ; where w is the moving window size, is the data after denoising, ensuring the smoothness of the environmental parameter data, and k is the time index within the moving window; S22. Detect outliers in the denoised environmental parameter data set, determine outliers according to the upper and lower limit ranges of each parameter and repair the outliers by interpolation; S23. Convert the denoised and repaired environmental parameter data into a time series matrix in a unified format: ; S24. Process the environmental parameter data of each partition according to the standardization formula to eliminate the influence of different parameter dimensions, and output the standardized environmental parameter data set .

[0010] Optionally, the specific steps of S3 include: S31. According to the type, growth stage of the crops in the greenhouse and the target yield requirement, combined with the pre-processed environmental parameter data set, obtain the input parameters of the partition priority model, including the crop health index , regional photosynthesis efficiency , regional resource utilization rate and partition area ratio , where i represents the partition number; S32. Based on the fuzzy logic reasoning results formed by leaf temperature, stomatal aperture and photosynthesis rate, obtain the real-time crop health index : ; where, represents the time variable, and are the start and end time points for collecting crop health data respectively, respectively represent the membership function mappings for leaf temperature, stomatal aperture, and photosynthesis rate, represents the T-norm operation, which is used to aggregate multiple health indicators, is at time the fuzzy logic inference output; S33. For the light intensity and concentration in partition i, quantify the photosynthesis efficiency: ; where t is the sampling time point, and are the start and end times for photosynthesis observation respectively, and respectively represent the membership degrees of light intensity and concentration, is the aggregation function, which is used to synthesize the membership degrees of light intensity and concentration into the local photosynthesis efficiency and accumulate it in the time dimension; S34. According to the dynamic relationship between the actual absorption amount and supply amount of in partition i, calculate the regional resource utilization rate : ; where n is the total number of samplings, represents the amount of actually absorbed by the crop in partition i at the t-th sampling moment, represents the total amount of supplied in partition i at the t-th sampling moment. When the difference between the actual absorption amount and supply amount is greater than the threshold, logarithmic correction can highlight resource waste or insufficient supply; S35. Based on the crop health index, regional photosynthesis efficiency, regional resource utilization rate, and partition area ratio, calculate the partition priority weight : ; where, is the weight factor.

[0011] Optionally, the specific steps of S4 include: S41. Based on the environmental parameter datasets of each partition and the partition priority weight W, define the input membership function set, which is used to describe the partition priority weight and Concentration of the fuzzy set, where for the fuzzy set of concentration below the threshold, the following piecewise exponential function is adopted: ; where u represents the normalized concentration, represents the segmentation point between the concentration below the threshold and other intervals, and are respectively the curve sensitivity coefficients in different intervals, is the high-order curve deformation coefficient; Describing the fuzzy membership degree corresponding to the concentration below the threshold in the greenhouse environment through the piecewise exponential function, meeting the modeling requirements for complex and variable environments; S42. Taking the regulation amount as the output variable of the fuzzy logic inference system, for the three fuzzy sets of small supply, medium supply and large supply, an adaptive deformation cosine function is defined to depict the fuzzy membership degree of the regulation amount: ; ; where z represents the output regulation amount, are respectively the adjustment parameters for controlling the cosine wave period, is the phase offset, used to distinguish intervals of different supply intensities; S43. Under the input dimensions of the partition priority weight and the normalized concentration , combining the membership degrees of other normalized environmental parameters in the greenhouse, multiple fuzzy rules are established. For each fuzzy rule, the following rule activation intensity is defined: ; where, represents the activation intensity of the l-th fuzzy rule, are respectively the fuzzy sets set for the partition priority weight, the normalized concentration, temperature and humidity in this rule, is the multi-element T-norm operation, used to combine the membership degrees of each input membership function. When is larger, the contribution weight of the rule in this round of reasoning is greater; S44. Aggregate all the rule activation intensities and the corresponding output membership degrees , and calculate the Regulation quantity: ; where L is the total number of fuzzy rules, is the priority coefficient of the l-th rule, which is set or adaptively updated according to the greenhouse operation experience, is the membership function corresponding to the output situation of the l-th rule, and z represents the achievable regulation quantity value.

[0012] Optionally, the specific steps of S5 include: S51. Input the pre-processed environmental parameter datasets and partition priority weights of each partition into the fuzzy logic inference system, and use the membership function to map the input variables into the corresponding fuzzy set membership degree matrix to quantitatively describe the current environmental state and priority weight characteristics of different partitions; S52. Perform inference operations on the input membership degree matrix based on the rule base stored in the fuzzy logic inference system, and generate a fuzzy output of the partition target concentration range. Through the joint activation between rules, the target range dynamically reflects the real-time environmental changes and partition requirements; S53. Convert the output result of the fuzzy rule inference into an actual value through the defuzzification method, and calculate the specific concentration regulation quantity of each partition; S54. Combine and output the target concentration range and regulation quantity of each partition according to the defuzzification result, and at the same time adjust the boundary of the regulation range in combination with the real-time environmental parameters to provide a dynamically adaptive supply strategy, so that the concentration distribution of each partition meets the crop growth requirements.

[0013] Optionally, the specific steps of S7 include: S71. Through the sensor network deployed in each partition of the greenhouse, collect the current environmental parameter dataset in real time, and compare the collected real-time data with the target concentration range to determine whether there is a deviation. If the actual concentration does not meet , that is , then it is determined that the concentration of partition i has a deviation; S72. For each partition, compare the actual concentration with the central value of the target range, and calculate the concentration deviation value : ; Among them, the target central value is the average value of the target concentration range, which is used to evaluate the insufficiency or excess of the supply in each partition; S73. Take the deviation value and the environmental parameter data collected in real time as inputs, restart the fuzzy logic inference system, perform correction calculations on the deviation through the established rule base, and output the updated concentration regulation amount; S74. Generate a new control signal according to the regulation amount obtained from the re-inference, and send it to the supply equipment, air flow regulating device or ventilation system in each partition, and execute the concentration adjustment strategy to correct the concentration deviation within the partition; S75. After the adjustment is executed, continuously collect real-time data through the sensor network to monitor the actual concentration in the partition. If, after multiple regulation cycles, the actual concentration concentration meets the target range , it is determined that the concentration in this partition reaches a stable state; if not, repeat steps S72 to S74 until all partitions are stable.

[0014] The beneficial effects of the present invention are as follows: (1) By constructing a partition priority model, the present invention comprehensively considers the crop health index, photosynthesis efficiency, resource utilization rate, and partition area ratio, dynamically calculates the priority weights of each partition, realizes the precise allocation of the supply. Through dynamic priority calculation, it ensures that high-priority partitions can obtain sufficient supply in a timely manner, while avoiding resource waste in low-demand areas, and significantly improving the overall resource utilization efficiency.

[0015] (2) The present invention inputs the environmental parameter data collected by the real-time sensor network into the fuzzy logic inference system, combines the rule base to dynamically calculate the target range and regulation amount of the concentration. Through the real-time monitoring and deviation value feedback mechanism, it can quickly respond to the complex changes in the environment and adjust the regulation signal, forming a closed-loop control system. The response speed in a dynamic environment is increased by more than 30%, and it effectively avoids the phenomena of over-supply or under-supply caused by response lag, and significantly reduces the operating cost.

[0016] (3)The present invention designs a flexible multi-dimensional membership function for precisely describing input parameters, and combines the centroid defuzzification method with rule weights to accurately calculate the regulation quantity. Through the piecewise exponential function and the adaptive cosine function non-linear model, the complex characteristics of environmental parameters can be more accurately reflected, improving the adaptability to the crop growth environment. Description of the Drawings

[0017] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention, and do not constitute a limitation to the present invention. In the drawings: Figure 1 A greenhouse concentration zoning regulation method based on fuzzy logic reasoning proposed by the present invention Flowchart of the method. Detailed Description of the Preferred Embodiments

[0018] Now, the present invention will be further described in detail with reference to the drawings. These drawings are all simplified schematic diagrams, only showing the basic structure of the present invention in a schematic way, so they only show the components related to the present invention.

[0019] Refer to Figure 1 , a greenhouse concentration zoning regulation method based on fuzzy logic reasoning, includes the following steps: S1. Real-time collect the environmental parameter datasets corresponding to each zone through the sensor network deployed inside the greenhouse; S2. Perform data preprocessing on the collected environmental parameter datasets of each zone; S3. Construct a greenhouse zone priority model according to the types, growth stages of the crops in the greenhouse and the target yield requirements. The zone priority model determines the priority weights of the zones by evaluating the importance and response sensitivity of the crops in each zone; S4. Based on the greenhouse zone priority model and the preprocessed environmental parameter datasets of each zone, combine with the fuzzy logic inference system to construct a concentration regulation rule base, The concentration regulation rule base includes the definition of the membership function, the setting of the fuzzy rules, and the correlation relationship between the input and output variables; S5. Use the fuzzy logic inference system to calculate the target concentration range and regulation quantity of each zone according to the real-time environmental parameter datasets of each zone and the priority weights of the greenhouse zone priority model; S6. Convert the target concentration range and regulation quantity of each zone into executable control signals, and transmit them to the zone executors inside the greenhouse A supply device, an air flow regulating device, and a ventilation system to perform concentration zone regulation and control; S7. Real-time monitor the environmental parameter data of each zone through a sensor network, and evaluate whether the actual concentration reaches the target range. If the actual concentration deviates from the target range, restart the fuzzy logic inference system for calculation based on the deviation value, and update the control signal until the target range is stable.

[0020] In this embodiment, the specific steps of S1 include: S11. Deploy multi-type sensor nodes in each zone through the sensor network inside the greenhouse to collect the environmental parameter data of each zone in real time. The environmental parameter data includes concentration , temperature , humidity , light intensity , air flow distribution characteristics , and plant growth stage , where i represents the zone number and j represents the time series index; S12. Construct the environmental parameter data collected from each zone into an environmental parameter data set D: ; where m represents the total number of zones in the greenhouse and n represents the total number of time series sampling points.

[0021] In this embodiment, the specific steps of S2 include: S21. Denoise each element in the collected environmental parameter data set D, and calculate the denoised value using the moving average filtering algorithm: ; where w is the moving window size, is the data after denoising, ensuring the smoothness of the environmental parameter data, and k is the time index within the moving window; S22. Detect outliers in the denoised environmental parameter data set, determine outliers according to the upper and lower limit ranges of each parameter , and repair the outliers by interpolation; S23. Convert the denoised and repaired environmental parameter data into a time series matrix in a unified format: ; S24. Process the environmental parameter data of each zone according to the standardization formula to eliminate the influence of different parameter dimensions, and output the environmental parameter data set after the standardization process is completed.

[0022] In this embodiment, the specific steps of S3 include: S31. According to the types, growth stages of the crops in the greenhouse and the target yield requirements, combined with the pre-processed environmental parameter dataset, obtain the input parameters of the zoning priority model, including the crop health index , regional photosynthesis efficiency , regional resource utilization rate and the zoning area ratio , where i represents the zoning number; S32. Based on the fuzzy logic inference results formed by the leaf temperature, stomatal aperture and photosynthesis rate, obtain the real-time crop health index : ; wherein, represents the time variable, and are the start and end time points of crop health data collection respectively, respectively represent the membership function mappings for the leaf temperature, stomatal aperture and photosynthesis rate, represents the T-norm operation for aggregating multiple health indicators, is the fuzzy logic inference output at time ; S33. Quantify the photosynthesis efficiency for the light intensity and concentration in zone i: ; where t is the sampling time point, and are the start and end times of photosynthesis observation respectively, and respectively represent the membership degrees of the light intensity and concentration, is the aggregation function for synthesizing the membership degrees of the light intensity and concentration into the local photosynthesis efficiency and performing accumulation in the time dimension; S34. Calculate the regional resource utilization rate according to the dynamic relationship between the actual absorption amount and supply amount of in zone i: ; where n is the total number of samplings, represents the actual absorption amount of by the crops in zone i at the t-th sampling moment, Indicating the amount supplied within partition i at the t-th sampling moment When the difference between the actual absorption amount and the supply amount is greater than the threshold, logarithmic correction can highlight resource waste or insufficient supply; S35. Calculate the partition priority weight based on the crop health index, regional photosynthesis efficiency, regional resource utilization rate, and partition area ratio : ; Among them, is the weight factor.

[0023] In this embodiment, the specific steps of S4 include: S41. Based on the environmental parameter datasets of each partition and the partition priority weight W, define the input membership function set to describe the fuzzy sets of the partition priority weight and concentration Among them, for the fuzzy set where the concentration is lower than the threshold, the following piecewise exponential function is adopted: ; Among them, u represents the standardized concentration, represents the division point between the interval lower than the threshold and other intervals, and are respectively the curve sensitivity coefficients in different intervals, is the high-order curve deformation coefficient; Describe the fuzzy membership degree corresponding to the concentration lower than the threshold in the greenhouse environment through the piecewise exponential function, meeting the modeling requirements for complex and variable environments; S42. Take the regulation amount as the output variable of the fuzzy logic inference system. For the three fuzzy sets of small supply, medium supply, and large supply, define an adaptive deformation cosine function to depict the fuzzy membership degree of the regulation amount: ; Among them, z represents the output regulation amount, are respectively the adjustment parameters for controlling the cosine wave period, is the phase offset for distinguishing intervals of different supply intensities; S43. Among the partition priority weight and the standardized concentration Under the input dimension, by integrating the membership degrees of other standardized environmental parameters in the greenhouse, multiple fuzzy rules are established. For each fuzzy rule, the following rule activation intensity is defined: ; where, represents the activation intensity of the l-th fuzzy rule, are the fuzzy sets respectively set for the partition priority weight, standardized concentration, temperature and humidity of this rule, is the multi-variable T-norm operation, which is used to integrate the membership degrees of each input membership function. When is larger, the contribution weight of the rule in this round of reasoning is greater; S44. Aggregate the activation intensities of all rules and the corresponding output membership degrees , and calculate the regulation amount of each partition through the centroid defuzzification formula with rule weights: ; where, L is the total number of fuzzy rules, is the priority coefficient of the l-th rule, which is set or adaptively updated according to the greenhouse operation experience, is the membership function corresponding to the output situation of the l-th rule, and z represents the value of the achievable regulation amount.

[0024] In this embodiment, the specific steps of S5 include: S51. Input the environmental parameter datasets of each partition and the partition priority weights after data preprocessing into the fuzzy logic inference system, and use the membership function to map the input variables into the corresponding fuzzy set membership degree matrix to quantitatively describe the current environmental state and priority weight characteristics of different partitions; S52. Based on the rule base stored in the fuzzy logic inference system, perform inference operations on the input membership degree matrix, and generate a fuzzy output of the partition target concentration range by integrating the activation intensities of multiple rules. Through the joint activation between rules, the target range dynamically reflects the real-time environmental changes and partition requirements; S53. Convert the output result of the fuzzy rule inference into an actual value through the defuzzification method, and calculate the specific concentration regulation amount of each partition; S54. Combine and output the target concentration range and regulation amount of each partition according to the defuzzification result, and at the same time adjust the boundary of the regulation range in combination with the real-time environmental parameters to provide a dynamically adaptive supply strategy, so that the concentration distribution of each partition meets the crop growth requirements.

[0025] In this embodiment, the specific steps of S7 include: S71. Through the sensor network deployed in each partition of the greenhouse, the current environmental parameter dataset is collected in real time, and the collected real-time data is compared with the target concentration range to determine whether there is a deviation. If the actual concentration does not meet , that is , it is determined that the concentration in partition i has a deviation; S72. For each partition, the actual concentration is compared with the central value of the target range to calculate the concentration deviation value : ; Among them, the target central value is the average value of the target concentration range, which is used to evaluate the shortage or excess of the supply in the partition ; S73. Taking the deviation value and the environmental parameter data collected in real time as inputs, restart the fuzzy logic inference system, and perform correction calculations on the deviation through the established rule base to output the updated concentration regulation amount; S74. Generate a new control signal according to the regulation amount obtained from the re-inference, and send it to the supply equipment, air flow adjustment device or ventilation system in each partition to execute the concentration adjustment strategy to correct the concentration deviation in the partition; S75. After the adjustment is executed, continuously collect real-time data through the sensor network to monitor the actual concentration in the partition. If after multiple regulation cycles, the actual concentration meets the target range , it is determined that the concentration in this partition has reached a stable state; if not, repeat steps S72 to S74 until all partitions are stable.

[0026] Example 1: A certain agricultural technology company carried out a project on Research on precise regulation of concentration zones. The greenhouse has an area of 5000 square meters and is divided into 25 zones, each with an area of approximately 200 square meters, where high-value cash crops such as tomatoes, cucumbers, and peppers are planted respectively. A high-precision sensor network is equipped in the greenhouse to collect environmental parameters of each zone in real time, including concentration, temperature, humidity, and air flow distribution information. Due to differences in crop types, growth stages, and environmental conditions among zones, the demand for shows significant spatial non-uniformity. Traditional overall regulation methods are difficult to meet this differentiated demand, resulting in excess in some areas while shortages in other areas, which not only wastes resources but also affects crop growth efficiency.

[0027] To solve the above problems, the greenhouse operator introduced the greenhouse concentration zone regulation method based on fuzzy logic reasoning of the present invention, aiming to achieve precise regulation of concentration within the zone through a dynamic priority model, fuzzy logic reasoning, and real-time monitoring closed-loop control.

[0028] One day, the concentration supply in the tomato planting area of the greenhouse is insufficient, while the concentration in the cucumber planting area is too high, resulting in a decrease in the photosynthesis efficiency of tomatoes and curling of cucumber leaves. Traditional overall regulation methods cannot adjust for a single zone and can only solve the problem in the tomato area by increasing the overall supply amount, but this further exacerbates the concentration excess in the cucumber area.

[0029] The management system of the greenhouse activates the method of the present invention. First, the following data is collected through the sensor network: the real-time concentration in the tomato area is 350 ppm, lower than its optimal concentration range of 400 - 450 ppm; the real-time concentration in the cucumber area is 550 ppm, higher than its optimal concentration range of 400 - 500 ppm; the concentration in the pepper area is 450 ppm and is in a stable state. In addition, the sensor also records the temperature (25 - 28 °C), humidity (60% - 70%), and air flow distribution characteristics of each area.

[0030] According to the method of the present invention, the system first preprocesses the environmental parameter data of each zone collected, removes noise and outliers, and uniformly standardizes it into a format suitable for fuzzy logic reasoning. Subsequently, the system dynamically calculates the priority weights of each zone based on the crop health index, photosynthesis efficiency, resource utilization rate, and area weight of the tomato, cucumber, and pepper areas. The calculation results show that the priority weight of the tomato area is 0.7, the cucumber area is 0.5, and the pepper area is 0.4.

[0031] Next, the system inputs the preprocessed data and priority weights into the fuzzy logic inference system, and performs inference calculations through the established rule base. The inference results show that the supply in the tomato area needs to be increased by 100 ppm, the supply in the cucumber area needs to be reduced by 50 ppm, while the pepper area does not need adjustment. Based on this, the system generates a control signal to instruct the supply equipment and ventilation system to adjust the tomato and cucumber areas respectively.

[0032] Within 10 minutes after the adjustment is executed, the sensor network continuously collects real-time data and performs closed-loop control to ensure that the concentration in each area gradually approaches the target range. Finally, the concentration in the tomato area stabilizes at 420 ppm, the cucumber area stabilizes at 480 ppm, and the pepper area remains at 450 ppm. All areas have reached the optimal concentration range.

[0033] To verify the effectiveness of the method of the present invention, the research team compared the effects of using the traditional overall regulation method and the method of the present invention. Under the same environmental conditions, for the requirements of the crops in 25 partitions, the experimental data are as follows: Traditional method: The overall supply is 1000 ppm, the actual concentration in the tomato area is 370 ppm, the actual concentration in the cucumber area is 580 ppm, and the actual concentration in the pepper area is 460 ppm; the average photosynthesis efficiency is 62%, and the resource utilization rate is 68%.

[0034] Method of the present invention: The total supply is 850 ppm, the actual concentration in the tomato area is 420 ppm, the actual concentration in the cucumber area is 480 ppm, and the actual concentration in the pepper area is 450 ppm; the average photosynthesis efficiency is 85%, and the resource utilization rate is 92%.

[0035] It can be seen from the data that due to the inability to achieve partition regulation, the traditional method leads to resource waste and local concentration imbalance, affecting the overall photosynthesis efficiency. However, the method of the present invention through precise partition control not only reduces the total supply, but also significantly improves the resource utilization rate and photosynthesis efficiency.

[0036] This embodiment verifies the feasibility and superiority of the method of the present invention in a complex greenhouse environment. Through the application of the dynamic priority model and fuzzy logic inference, the method of the present invention realizes the precise partition regulation of the greenhouse concentration, effectively solves the problems of resource waste and uneven regional demand under the traditional overall regulation method, and provides technical support for the efficient operation of intelligent greenhouses.

[0037] Through the construction of a partition priority model, the present invention comprehensively considers the crop health index, photosynthesis efficiency, resource utilization rate, and partition area ratio, dynamically calculates the priority weights of each partition, and realizes the precise allocation of supply. Through dynamic priority calculation, it ensures that high-priority partitions can obtain sufficient supply in a timely manner, while avoiding resource waste in low-demand areas, and significantly improving the overall resource utilization efficiency.

[0038] The present invention inputs the environmental parameter data collected by the real-time sensor network into the fuzzy logic inference system, and dynamically calculates the target range and regulation amount of the concentration in combination with the rule base. Through the real-time monitoring and deviation value feedback mechanism, it can quickly respond to the complex changes in the environment and adjust the regulation signal, forming a closed-loop control system. The response speed in the dynamic environment is increased by more than 30%, and it effectively avoids the phenomena of over-supply or under-supply caused by response lag, and significantly reduces the operating cost.

[0039] The present invention designs a flexible multi-dimensional membership function for the refined description of input parameters, and realizes the precise calculation of the regulation amount in combination with the centroid defuzzification method with rule weights. Through the piecewise exponential function and the adaptive cosine function nonlinear model, it can more accurately reflect the complex characteristics of environmental parameters and improve the adaptability to the crop growth environment.

[0040] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. A greenhouse based on fuzzy logic reasoning The concentration zoning control method is characterized in that: The steps include: S1. Collect the environmental parameter data sets corresponding to each partition in real time through the sensor network deployed inside the greenhouse; S2. Preprocessing the collected environmental parameter data sets for each partition; S3. A greenhouse zoning priority model is constructed based on the types, growth stages, and target yield requirements of the crops in the greenhouse. The zoning priority model evaluates the importance and Response sensitivity, determine the priority weight of the partition; S4. Based on the greenhouse partition priority model and the preprocessed environmental parameter data set of each partition, combined with the fuzzy logic reasoning system Concentration control rule base, The concentration control rule base includes the definition of membership function, the setting of fuzzy rules and the relationship between input and output variables; S5. Using the fuzzy logic reasoning system, the target of each partition is calculated based on the real-time environmental parameter data set of each partition and the priority weight of the greenhouse partition priority model. Concentration range and control amount; S6. Set the target of each partition The concentration range and control amount are converted into executable control signals and transmitted to the partitions in the greenhouse for execution. Supply equipment, air conditioning equipment and ventilation systems to perform Concentration zoning control; S7. Use the sensor network to monitor the environmental parameter data of each partition in real time and evaluate the actual Whether the concentration reaches the target range, if the actual If the concentration deviates from the target range, the fuzzy logic inference system is restarted to perform calculations based on the deviation value and the control signal is updated until the target range is stabilized.

2. A greenhouse based on fuzzy logic reasoning according to claim 1 The concentration zoning control method is characterized in that: The specific steps of S1 include: S11. Multiple types of sensor nodes are deployed in each partition through the sensor network inside the greenhouse to collect environmental parameter data of each partition in real time. The environmental parameter data includes concentration ,temperature ,humidity , light intensity , air flow distribution characteristics and plant growth stages , where i represents the partition number and j represents the time series index; S12. Constructing the collected environmental parameter data of each partition into an environmental parameter data set D: ； Where m represents the total number of greenhouse partitions, and n represents the total number of time series sampling points.

3. A greenhouse based on fuzzy logic reasoning according to claim 1 The concentration zoning control method is characterized in that: The specific steps of S2 include: S21. For each element in the collected environmental parameter data set D Perform denoising and use the sliding average filter algorithm to calculate the denoising value: ； Among them, w is the sliding window size, is the denoised data to ensure the smoothness of the environmental parameter data, and k is the time index within the sliding window; S22. Perform outlier detection on the denoised environmental parameter data set, based on the upper and lower limits of each parameter Determine outliers and repair them through interpolation; S23. Convert the denoised and restored environmental parameter data into a time series matrix in a unified format: ； S24. Process the environmental parameter data of each partition according to the standardized formula to eliminate the influence of different parameter dimensions, and complete the standardized environmental parameter data set Output.

4. A greenhouse based on fuzzy logic reasoning according to claim 1 The concentration zoning control method is characterized in that: The specific steps of S3 include: S31. Based on the types, growth stages and target yield requirements of crops in the greenhouse, combined with the environmental parameter data set after data preprocessing, obtain the input parameters of the zoning priority model, including the crop health index , regional photosynthesis efficiency , Regional resource utilization and area ratio of the partition , where i represents the partition number; S32. Obtain real-time crop health index based on fuzzy logic reasoning results formed by leaf temperature, stomatal aperture and photosynthesis rate : ； in, represents the time variable, and are the start and end time points of crop health data collection, They represent the membership function mapping for leaf temperature, stomatal aperture and photosynthesis rate respectively. Represents the T-norm operation, which is used to aggregate multiple health indicators. For in time The fuzzy logic reasoning output when ; S33. Light intensity for partition i and concentration Quantify photosynthetic efficiency: ； Where t is the sampling time point, and are the start and end times of photosynthesis observation, and Represents the light intensity and The degree of membership of the concentration, is an aggregation function that combines light intensity with The degree of membership of the concentration is integrated into the local photosynthesis efficiency and accumulated in the time dimension; S34. According to the partition i The dynamic relationship between the actual absorption and supply, calculate the regional resource utilization : ； Where n is the total number of sampling times, represents the actual absorption of crops in partition i at the tth sampling time quantity, represents the supply in partition i at the tth sampling time Total amount: when the difference between actual absorption and supply is greater than the threshold, logarithmic correction can be used to highlight resource waste or insufficient supply; S35. Calculate the priority weight of each zone based on the crop health index, regional photosynthesis efficiency, regional resource utilization rate, and zone area ratio. : ； in, is the weight factor.

5. A greenhouse based on fuzzy logic reasoning according to claim 1 The concentration zoning control method is characterized in that: The specific steps of S4 include: S41. Based on the environmental parameter data set of each zone And the partition priority weight W, define the input membership function set to describe the partition priority weight and concentration Fuzzy set of The fuzzy set with concentration below the threshold uses the following piecewise exponential function: ； Among them, u represents the standardization concentration, Indicates the split point with other intervals below the threshold, and are the curve sensitivity coefficients in different intervals, is the deformation coefficient of the high-order curve; The piecewise exponential function describes the The fuzzy membership corresponding to the concentration satisfies the complex and changing Modeling requirements for the environment; S42. Control volume As the output variable of the fuzzy logic reasoning system, the cosine function of adaptive deformation is defined for the three fuzzy sets of small supply, medium supply and large supply to characterize Fuzzy membership of the control quantity: ； Among them, z represents the output control amount, are the adjustment parameters for controlling the cosine wave period, is the phase offset, which is used to distinguish the intervals with different supply intensities; S43. Priority weight in partition and Standardization concentration Under the input dimension, the membership of other standardized environmental parameters in the greenhouse is integrated to establish multiple fuzzy rules. For each fuzzy rule, the following rule activation strength is defined: ； in, represents the activation strength of the lth fuzzy rule, The rule is respectively for the partition priority weight and standardization Fuzzy set set by concentration, temperature and humidity, is a multivariate T-norm operation, which is used to synthesize the membership of each input membership function. The larger it is, the greater the contribution weight of the rule in this round of reasoning; S44. Activation strength for all rules And the corresponding output membership Aggregation is performed and the centroid defuzzification formula with rule weights is used to calculate the Control volume: ； Where L is the total number of fuzzy rules, is the priority coefficient of the first rule, which is set or adaptively updated according to greenhouse operation experience. is the membership function corresponding to the output of rule l, z represents the achievable Control quantity value.

6. A greenhouse based on fuzzy logic reasoning according to claim 1 The concentration zoning control method is characterized in that: The specific steps of S5 include: S51. Input the pre-processed environmental parameter data sets and priority weights of each partition into the fuzzy logic reasoning system, map the input variables into the corresponding fuzzy set membership matrix using the membership function, and quantitatively describe the current environmental status and priority weight characteristics of different partitions; S52. Perform reasoning operations on the input membership matrix based on the rule base stored in the fuzzy logic reasoning system, and generate partition targets by integrating the activation strength of multiple rules Fuzzy output of concentration range, through joint activation between rules, the target range dynamically reflects real-time environmental changes and zoning requirements; S53. The output results of fuzzy rule reasoning are converted into actual values ​​through defuzzification method, and the specific Concentration control amount; S54. According to the defuzzification results, the target of each partition The concentration range and control amount are combined and output, and the boundaries of the control range are adjusted in combination with real-time environmental parameters to provide a dynamic and adaptive supply strategy to ensure that the concentration of each partition is consistent with that of the control amount. The concentration distribution meets the needs of crop growth.

7. A greenhouse based on fuzzy logic reasoning according to claim 1 The concentration zoning control method is characterized in that: The specific steps of S7 include: S71. Through the sensor network deployed in each partition of the greenhouse, the current environmental parameter data set is collected in real time, and the collected real-time data is compared with the target Compare the concentration range to determine whether there is a deviation. concentration Dissatisfied ,Right now , then determine the partition i Concentration deviations; S72. For each partition, the actual concentration The center value of the target range Compare and calculate the concentration deviation value : ； The target center value is the average value of the target concentration range and is used to evaluate the partition shortage or excess of supply; S73. Set the deviation value The real-time collected environmental parameter data is used as input, the fuzzy logic reasoning system is restarted, the deviation is corrected and calculated through the established rule base, and the updated Concentration control amount; S74. Based on the re-reasoning The control quantity generates a new control signal and sends it to each partition Supply equipment, air conditioning devices or ventilation systems, perform concentration adjustment strategies to correct concentration deviations within the partitions; S75. After the adjustment is executed, continue to collect real-time data through the sensor network to monitor the actual The concentration is monitored. If the actual concentration Meet target range , then determine the partition The concentration reaches a stable state; if not, steps S72 to S74 are repeated until all partitions are stable.

Images