Intelligent kiwi fruit irrigation method and system adopting distributed optimization control

Abstract

The invention provides an intelligent kiwi fruit irrigation method and system adopting distributed optimization control, and belongs to the field of intelligent control. The method comprises the steps of obtaining weather prediction data, initial irrigation parameters and expected growth vigor of kiwi fruits of a target area in a future time zone; environment zoning and kiwi fruit growth zoning are carried out to obtain a first zoning result and a second zoning result, and the first zoning result and the second zoning result intersect to obtain a target area zoning result; predicting the growth vigor to obtain a kiwi fruit growth vigor prediction result; and extracting a partition in which the kiwi fruit growth prediction result is inconsistent with the expected kiwi fruit growth, executing irrigation parameter optimization, obtaining distributed irrigation parameters, and executing kiwi fruit distributed irrigation control. The technical problems that in the prior art, kiwi fruit irrigation is low in intelligent degree, and precise irrigation cannot be conducted according to different regional environments and growth differences are solved, and the technical effects that the kiwi fruit irrigation intelligent level is improved, and distributed precise irrigation based on the regional environments and the growth differences is achieved are achieved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control, and in particular to a kiwifruit intelligent irrigation method and system using distributed optimization control. Background Art

[0002] Kiwifruit has a fleshy root system primarily concentrated in the shallow 20-40 cm soil layer. This makes it sensitive to changes in soil moisture and structure, requiring precise irrigation management to ensure normal growth and high-quality fruit. Currently, kiwifruit is irrigated using a traditional, whole-area irrigation model, with fixed irrigation according to growing seasons. When the weather changes, growers increase irregular irrigation based on experience and the tree's growth.

[0003] However, this traditional irrigation method lacks intelligence and lacks the ability to predict and analyze environmental factors and tree growth, making it difficult to achieve precise water management. Furthermore, it relies heavily on the expertise of growers, with irrigation decisions primarily based on experience and lacking objective data support. Furthermore, traditional methods are unable to differentiate irrigation based on regional environmental conditions and kiwifruit growth patterns, leading to over-irrigation or under-irrigation in some areas, impacting kiwifruit growth and quality. Summary of the Invention

[0004] The present invention aims to solve the technical problems in the existing technology of low intelligence level of kiwifruit irrigation and inability to perform precise irrigation according to the environment and growth differences of different regions, and provides a kiwifruit intelligent irrigation method and system using distributed optimization control to solve the problem.

[0005] The technical solution of the present invention to solve the above technical problems is as follows: In a first aspect, the present invention provides a kiwifruit smart irrigation method using distributed optimization control, which is applied to a kiwifruit distributed irrigation controller. The method comprises: obtaining weather forecast data, initial irrigation parameters, and expected kiwifruit growth for a target area in a future time zone; performing environmental zoning on the target area to obtain a first zoning result; performing kiwifruit growth zoning on the target area to obtain a second zoning result; intersecting the first zoning result with the second zoning result to obtain a target area zoning result; traversing the target area zoning results, extracting kiwifruit monitored growth in each zoning area, and performing growth prediction based on the weather forecast data and the initial irrigation parameters to obtain a kiwifruit growth prediction result, wherein the kiwifruit growth prediction result corresponds one-to-one with the target area zoning result; extracting kiwifruit growth prediction results for zoning areas where the kiwifruit growth prediction results are inconsistent with the expected kiwifruit growth, performing irrigation parameter optimization to obtain distributed irrigation parameters, and performing kiwifruit distributed irrigation control.

[0006] In a second aspect, the present invention provides a kiwifruit smart irrigation system using distributed optimization control, which is applied to a kiwifruit distributed irrigation controller. The system includes: a data acquisition module for obtaining weather forecast data, initial irrigation parameters, and expected kiwifruit growth in a target area for a future time zone; a dual partitioning module for performing environmental partitioning on the target area to obtain a first partitioning result, and performing kiwifruit growth partitioning on the target area to obtain a second partitioning result; a partition intersection module for intersecting the first and second partitioning results to obtain a target area partitioning result; a growth prediction module for traversing the target area partitioning results, extracting the monitored growth of kiwifruit in each partition, and performing growth prediction based on the weather forecast data and the initial irrigation parameters to obtain a kiwifruit growth prediction result, wherein the kiwifruit growth prediction result corresponds one-to-one with the target area partitioning result; and an optimization control module for extracting partitions where the kiwifruit growth prediction result is inconsistent with the expected kiwifruit growth, performing irrigation parameter optimization, obtaining distributed irrigation parameters, and executing kiwifruit distributed irrigation control.

[0007] The beneficial effects of the present invention are: Obtain weather forecast data, initial irrigation parameters and expected kiwifruit growth in the future time zone of the target area to provide basic data support for subsequent distributed irrigation control; perform environmental zoning on the target area to obtain the first zoning result, and perform kiwifruit growth zoning on the target area to obtain the second zoning result, and realize fine division of the target area through double partitioning; intersect the first zoning result and the second zoning result to obtain the target area zoning result, ensuring that each zoning takes into account both environmental factors and kiwifruit growth differences; traverse the target area zoning results, extract the kiwifruit monitoring growth in the zoning, combine the weather forecast data and the initial irrigation parameters, perform growth prediction, and obtain the kiwifruit growth prediction result, wherein the kiwifruit growth prediction result corresponds one-to-one with the target area zoning result, realizing the prediction of the future growth of each zoning; extract the kiwifruit growth prediction result and the expected kiwifruit growth, perform irrigation parameter optimization, obtain distributed irrigation parameters, execute kiwifruit distributed irrigation control, and realize accurate distributed optimization control of the zoning that needs to be adjusted.

[0008] Through the above technical solution, intelligent irrigation of kiwifruit based on distributed optimization control is realized, which changes the traditional fixed mode of overall irrigation of the entire area and adopts non-periodic and zoned distributed irrigation methods, thereby improving the intelligence level of kiwifruit irrigation. It can perform differentiated and precise irrigation control according to the environmental conditions of different regions and the growth differences of kiwifruit, effectively solving the technical problems of low intelligence and inability to accurately irrigate in existing technologies. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1A schematic diagram of a process flow of a kiwifruit smart irrigation method using distributed optimization control provided by the present invention; Figure 2 This is a structural schematic diagram of a kiwifruit intelligent irrigation system using distributed optimization control provided by the present invention.

[0010] In the accompanying drawings, the components represented by the reference numerals are as follows: Data acquisition module 11, dual partitioning module 12, partition intersection module 13, growth prediction module 14, optimization control module 15. DETAILED DESCRIPTION

[0011] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0012] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the specified features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0013] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or illustration". Any embodiment of the present invention described as "for example" is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed herein.

[0014] Example 1, as Figure 1 As shown, an embodiment of the present invention provides a kiwifruit smart irrigation method using distributed optimization control, which is applied to a kiwifruit distributed irrigation controller, including: S1. Obtain weather forecast data for the target area in the future time zone, initial irrigation parameters, and expected kiwifruit growth conditions.

[0015] Specifically, the target area refers to the kiwifruit planting area that requires smart irrigation control. This area can be a single orchard, multiple contiguous orchards, or a large-scale kiwifruit planting base. After the target area is determined, weather forecast data for the target time zone, initial irrigation parameters, and expected kiwifruit growth are obtained.

[0016] Weather forecast data for the future time zone includes, but is not limited to, meteorological factors such as rainfall, temperature, humidity, wind speed, and sunshine duration for the next 7-15 days. This weather forecast data is obtained through a data interface with the meteorological department or a third-party weather forecast service platform and stored in a time series format. The time resolution can be set to hourly or daily levels based on actual needs. Weather forecast data should be obtained covering the target area and its surrounding affected areas to ensure the accuracy and representativeness of the forecast.

[0017] Initial irrigation parameters are the basic irrigation control parameters set before the kiwifruit distributed irrigation controller is operational. These parameters, including irrigation frequency, duration, intensity, and water temperature, are directly uploaded and inputted via the user end. These parameters can be based on traditional experience or by referencing historical irrigation data from similar kiwifruit growing areas, providing foundational data support for subsequent optimization of irrigation parameters.

[0018] Expected kiwifruit growth is pre-set by the user based on planting objectives and cultivar characteristics. This includes expected values ​​for growth indicators such as leaf number, internode length, shoot diameter, leaf shape index, and fruit development. Expected growth should be tailored to the kiwifruit's growth cycle, with corresponding parameter ranges set for different growth stages. Expected growth data is stored in numerical form to facilitate subsequent quantitative comparison with predicted growth.

[0019] By obtaining weather forecast data for future time zones in the target area, initial irrigation parameters, and expected kiwifruit growth, a data foundation is laid for distributed irrigation control.

[0020] S2. Perform environmental zoning on the target area to obtain a first zoning result, and perform kiwifruit growth zoning on the target area to obtain a second zoning result.

[0021] Specifically, the target area was divided into environmental zoning and kiwifruit growth zoning, and the first zoning results and the second zoning results were obtained respectively.

[0022] First, the target area is environmentally zoned. Specifically, the target area is divided into several sub-areas with similar environments based on the differences in environmental attributes at different locations within the target area. Environmental zoning primarily considers key environmental factors that affect kiwifruit growth, including but not limited to soil moisture, soil temperature, soil pH, light intensity, altitude, slope, wind speed, and other environmental attributes. By performing cluster analysis on the distribution of monitored values ​​for these environmental attributes, areas with similar environmental characteristics are identified and merged to obtain the first zoning result. Each sub-area in the first zoning result has relatively uniform environmental conditions, providing an environmental foundation for subsequent precision irrigation control.

[0023] At the same time, the target area is zoned according to the growth of kiwifruit. Specifically, based on the differences in the actual growth conditions of kiwifruit in the target area, the target area is divided into several sub-areas with similar growth. The kiwifruit growth zoning considers key indicators reflecting the growth status of kiwifruit, including but not limited to growth parameters such as the number of leaves, internode length, new shoot thickness, leaf shape index, and fruit development status. The zoning process is the same as the environmental zoning process. By performing cluster analysis on the distribution information of kiwifruit growth monitoring values, areas with similar growth characteristics are identified and merged to obtain the second zoning result. Each zone in the second zoning result has a relatively consistent kiwifruit growth level, providing a growth basis for subsequent growth prediction and irrigation optimization.

[0024] By zoning the target area for environment and kiwifruit growth, we obtained the first zoning results and the second zoning results respectively, laying the foundation for subsequent comprehensive zoning.

[0025] S3. Intersect the first partitioning result and the second partitioning result to obtain a target area partitioning result.

[0026] Specifically, the first partitioning result (environmental partitioning result) and the second partitioning result (kiwifruit growth partitioning result) are spatially intersected to form a comprehensive target area partitioning result.

[0027] Based on the intersection of the first and second zoning results, a spatial overlay analysis was performed to geometrically intersect the boundaries of each environmental zone in the first zoning result with the boundaries of the kiwifruit growth zones in the second zoning result. During this intersection, overlapping areas of the two zoning results were identified, and new zone boundaries were generated based on this overlap. Each new zone generated by this intersection possessed relatively uniform environmental conditions and similar kiwifruit growth characteristics, forming a comprehensive zoning unit under the dual constraints of environment and growth.

[0028] The target area zoning results are the final zoning scheme formed after the intersection. Each zone in this target area zoning result has clear spatial boundaries and dual attribute characteristics. Specifically, each zone contains environmental parameters (such as soil moisture, soil temperature, and light intensity) and growth parameters (such as leaf number, internode length, and shoot diameter) for that area. These parameters are relatively consistent within a zone, but vary significantly between zones. The target area zoning results provide precise spatial units for subsequent zone growth prediction and personalized irrigation parameter optimization.

[0029] By intersecting the first partition results with the second partition results, the target area partition results that take into account both environmental conditions and kiwifruit growth are obtained, achieving a refined division of the target area.

[0030] S4. Traverse the target area partition results, extract the kiwifruit monitoring growth in the partition, combine the weather forecast data and the initial irrigation parameters, perform growth forecast, and obtain kiwifruit growth forecast results, wherein the kiwifruit growth forecast results correspond one-to-one to the target area partition results.

[0031] Specifically, by processing each partition in the target area partition results one by one, the kiwifruit monitoring growth data of each partition is extracted, and the growth prediction is carried out in combination with weather forecast data and initial irrigation parameters to form a kiwifruit growth prediction result.

[0032] First, each partition in the target region's partitioning results is accessed sequentially, and the same growth prediction process is performed for each partition. During the traversal process, the kiwifruit monitoring growth data for the current partition is extracted, including current values ​​for growth parameters such as leaf number, internode length, shoot diameter, leaf shape index, and fruit development. Then, based on the monitored growth data for the partitioned kiwifruit, weather forecast data, and initial irrigation parameters, the expected growth changes for the kiwifruit in that partition are predicted for the future. During the prediction process, the partition's environmental parameters and current growth parameters are combined with the weather forecast data and initial irrigation parameters. The prediction is then calculated using the target region's representative sample growth prediction model to derive the predicted kiwifruit growth value for that partition in the future. The predicted kiwifruit growth values ​​for each partition derived from the growth prediction are then summarized to form a kiwifruit growth prediction result. This kiwifruit growth prediction result corresponds one-to-one with the target region's partitioning results; that is, each partition has a corresponding growth prediction result, containing predicted values ​​for growth parameters such as leaf number, internode length, and shoot diameter for the future.

[0033] By traversing the zoning results of the target area and performing growth predictions, the kiwifruit growth prediction results corresponding to each zone were obtained, providing prediction data support for subsequent irrigation parameter optimization.

[0034] S5. Extract the partitions where the kiwifruit growth prediction result is inconsistent with the expected kiwifruit growth, perform irrigation parameter optimization, obtain distributed irrigation parameters, and perform kiwifruit distributed irrigation control.

[0035] Specifically, by comparing the predicted growth results of kiwifruit in each partition with the expected growth of kiwifruit, the partitions with substandard growth are identified, and irrigation parameter optimization is performed on these partitions to achieve distributed irrigation control.

[0036] First, the kiwifruit growth prediction results of each partition are compared with the expected kiwifruit growth one by one, and the kiwifruit growth deviation between the two is calculated. When the kiwifruit growth deviation is greater than or equal to the preset growth parameter deviation threshold, the partition is identified as a partition with inconsistent growth, and irrigation parameter optimization is required. When the kiwifruit growth deviation is less than the growth parameter deviation threshold, the partition is identified as a partition with consistent growth, and the current irrigation parameters are maintained. Subsequently, the irrigation parameter optimization process is performed on the identified inconsistent growth partitions. The optimization process initializes multiple sets of irrigation parameters, calculates the growth prediction results corresponding to each set of irrigation parameters, and compares them with the expected kiwifruit growth. The process is iteratively executed and gradually converges to the optimal irrigation parameters. Afterwards, the corresponding optimal irrigation parameters are determined for each inconsistent growth partition, and combined with the maintenance parameters of the consistent growth partition to form distributed irrigation parameters. The distributed irrigation parameters contain the personalized irrigation control parameters of each partition, and differentiated irrigation control operations are performed on the corresponding partitions based on the distributed irrigation parameters.

[0037] By extracting inconsistent growth zones and performing irrigation parameter optimization, targeted distributed irrigation parameters were obtained, and distributed irrigation control of kiwifruit based on growth prediction was realized.

[0038] Furthermore, the target area is partitioned to obtain a first partition result, including: S21, collecting the first environmental attribute monitoring value distribution information of the target area, performing cluster analysis, and obtaining the first environmental attribute partition result; S22, until the Nth environmental attribute monitoring value distribution information of the target area is collected, cluster analysis is performed, and the Nth environmental attribute partition result is obtained; S23: Intersect the first environmental attribute partitioning results up to the Nth environmental attribute partitioning result to obtain the first partitioning result.

[0039] In a preferred embodiment, when performing environmental zoning on the target area and obtaining the first zoning result, first, the first environmental attribute monitoring value distribution information of the target area is collected. The first environmental attribute can be any one of the environmental factors that affect the growth of kiwifruit, such as soil moisture, soil temperature, soil pH value, light intensity, altitude, slope, slope direction, wind speed, etc. The real-time monitoring value of the first environmental attribute at different spatial positions is obtained by an environmental monitoring sensor network arranged in the target area to form the first environmental attribute monitoring value distribution information. The first environmental attribute monitoring value distribution information is stored in the form of coordinate-value pairs, recording the spatial position of each monitoring point in the target area and the corresponding monitoring value of the first environmental attribute. Subsequently, a clustering analysis method such as a K-means clustering algorithm, a hierarchical clustering algorithm or a DBSCAN clustering algorithm is used to process the first environmental attribute monitoring value distribution information, classify spatial positions with similar environmental attribute monitoring values ​​into the same category, and form continuous partition boundaries based on the principle of spatial continuity, thereby obtaining the first environmental attribute partition result.

[0040] Then, following the same process, the monitoring value distribution information for the target area's second, third, and Nth environmental attributes is collected. Each environmental attribute is collected using a dedicated sensor, such as a soil moisture sensor, temperature sensor, pH meter, or light meter. Cluster analysis is performed on the monitoring value distribution information for each environmental attribute, using the same algorithm parameters to ensure consistency across all partitioning results. Cluster analysis of each environmental attribute generates a corresponding partitioning result, ultimately yielding partitioning results for the second, third, and Nth environmental attributes. N is the total number of environmental attributes that affect kiwifruit growth.

[0041] Subsequently, spatial intersection operations are performed on the first environmental attribute partitioning results until the Nth environmental attribute partitioning results. The intersection process uses GIS spatial analysis technology to perform superposition analysis on the geometric boundaries of each environmental attribute partitioning result. Specifically, the partition boundary of the first environmental attribute partitioning result is intersected with the partition boundary of the second environmental attribute partitioning result to generate a new partition boundary, and then the new partition boundary is intersected with the third environmental attribute partitioning result, and so on until the intersection with the Nth environmental attribute partitioning result is completed. After gradual intersection, the original environmental attribute partitions are re-divided to form new partition units. Each new partition unit inherits the partition characteristics of multiple environmental attributes at the same time, that is, multiple environmental parameters such as soil moisture, temperature, pH value, light intensity, etc. within each partition are relatively uniform. The first partition result finally formed contains several sub-areas with relatively uniform comprehensive environmental conditions.

[0042] Through stepwise cluster analysis and spatial intersection of multiple environmental attributes, the first zoning result that comprehensively considers multiple environmental factors was obtained. This result can accurately reflect the spatial distribution characteristics of environmental conditions in the target area and provide an accurate environmental zoning basis for subsequent growth prediction and irrigation control.

[0043] The process of zoning the target area for kiwifruit growth is the same as the environmental zoning process. First, the distribution information of the monitoring value of the first growth attribute of the target area is collected. The first growth attribute can be any one of the indicators reflecting the growth status of kiwifruit, such as the number of leaves, internode length, new shoot thickness, leaf shape index, fruit development status, etc. The monitoring value of the first growth attribute at different spatial positions is obtained through image recognition technology or special sensors to form spatial distribution information. Subsequently, the monitoring value distribution information is processed using a cluster analysis algorithm, and the spatial positions with similar growth attribute monitoring values ​​are classified into the same category to obtain the first growth attribute zoning result. Continue to collect the monitoring value distribution information of the second growth attribute, the third growth attribute, and the Mth growth attribute of the target area, and perform cluster analysis on each growth attribute respectively to obtain the second growth attribute zoning result, the third growth attribute zoning result, and the Mth growth attribute zoning result. Wherein, M is the total number of growth attributes reflecting the growth status of kiwifruit. Next, a spatial intersection operation is performed on the results of the first growth attribute partition through the Mth growth attribute partition. The geometric boundaries of each growth attribute partition are superimposed and analyzed, and new partition units are formed after a step-by-step intersection operation. Each new partition unit inherits the characteristics of multiple growth attribute partitions. That is, multiple growth parameters such as leaf number, internode length, and new shoot diameter are relatively uniform within each partition, ultimately obtaining the second partition result.

[0044] Furthermore, the first environmental attribute monitoring value distribution information of the target area is collected and cluster analysis is performed to obtain the first environmental attribute partition result, including: S211. Extracting an environmental attribute monitoring value at a first position and an environmental attribute monitoring value at a second position from the first environmental attribute monitoring value distribution information, wherein the first position and the second position are adjacent positions; S212. Calculate the environmental attribute monitoring value deviation between the first location environmental attribute monitoring value and the second location environmental attribute monitoring value; S213: When the environmental attribute monitoring value deviation is less than a first environmental attribute deviation threshold, merging the first location and the second location into a third location, and calculating an average of the environmental attribute monitoring value of the first location and the environmental attribute monitoring value of the second location, setting the average as the environmental attribute monitoring value of the third location, where the first environmental attribute deviation threshold is a user-preset value; S214: When the environmental attribute monitoring value deviation is greater than or equal to a first environmental attribute deviation threshold, the first position and the second position are regarded as different positions; S215: When the deviations of the environmental attribute monitoring values ​​of any two adjacent positions are greater than or equal to a first environmental attribute deviation threshold, output the first environmental attribute partitioning result.

[0045] In a preferred embodiment, when performing cluster analysis on the first environmental attribute monitoring value distribution information, iterative clustering based on neighborhood merging is adopted. By gradually comparing the differences in environmental attribute monitoring values ​​of adjacent locations, similar locations are merged into the same partition until further merging is impossible.

[0046] First, pairs of adjacent locations are extracted from the first environmental attribute monitoring value distribution information according to a predetermined traversal order. The first environmental attribute monitoring value distribution information is stored as spatial coordinates and corresponding monitoring values, and includes the location information and first environmental attribute values ​​of all monitoring points within the target area. Each pair of adjacent locations, namely the first location and the second location, is sequentially extracted, and the corresponding environmental attribute monitoring values ​​are obtained as comparison objects, namely, the first location environmental attribute monitoring value and the second location environmental attribute monitoring value.

[0047] Next, the difference between the environmental attribute monitoring value at the first location and the environmental attribute monitoring value at the second location is calculated to obtain the environmental attribute monitoring value deviation. Specifically, the environmental attribute monitoring value deviation is calculated using the absolute difference method: environmental attribute monitoring value deviation = |environmental attribute monitoring value at the first location - environmental attribute monitoring value at the second location|. This environmental attribute monitoring value deviation reflects the degree of difference in environmental conditions between adjacent locations and serves as an indicator for determining whether to merge locations.

[0048] Subsequently, the calculated environmental attribute monitoring value deviation is compared with the first environmental attribute deviation threshold. The first environmental attribute deviation threshold is a numerical parameter pre-set by the user based on the natural variation range of the specific environmental attribute, the measurement accuracy of the monitoring equipment, the desired partitioning precision, and the actual application requirements. When the environmental attribute monitoring value deviation is less than the first environmental attribute deviation threshold, it indicates that the environmental attributes of the two adjacent positions are highly similar and meet the merging conditions. At this time, a position merging operation is performed to geometrically merge the spatial areas of the first position and the second position to generate a new third position. At the same time, the arithmetic mean of the environmental attribute monitoring value of the first position and the environmental attribute monitoring value of the second position is calculated as the environmental attribute monitoring value of the third position, representing the overall environmental characteristics of the merged area. When the environmental attribute monitoring value deviation is greater than or equal to the first environmental attribute deviation threshold, the first position and the second position are regarded as different positions, and no merging operation is performed, indicating that the environmental attributes of the two adjacent positions are significantly different and should belong to different partitions. The independence of the two positions is maintained, and the next pair of adjacent positions is processed.

[0049] Continue to repeat the above comparison and merging process to traverse all possible adjacent position pairs. After completing a complete round of traversal, check whether there are still adjacent position pairs with deviations less than the threshold. If still exist, continue with the next round of comparison and merging operations. When the deviations of the environmental attribute monitoring values ​​of any two adjacent positions are greater than or equal to the first environmental attribute deviation threshold, it indicates that all positions that meet the merging conditions have been merged and the clustering process ends. At this time, the first environmental attribute partitioning result is output. The first environmental attribute partitioning result contains several spatially continuous partitions. The environmental attribute monitoring values ​​within each partition are relatively uniform, and there are obvious differences in environmental attributes between partitions.

[0050] Through cluster analysis based on the gradual merging of adjacent locations, the target area can be divided into different zones according to the spatial similarity of environmental attributes, providing an accurate environmental zoning basis for subsequent multi-attribute zone intersection and irrigation control.

[0051] Furthermore, the target area partition results are traversed, the growth of kiwifruit in each partition is extracted, and the growth prediction is performed in combination with the weather forecast data and the initial irrigation parameters to obtain the kiwifruit growth prediction result, including: S41. Obtaining the kiwifruit planting duration, accumulating the future time zones, and obtaining a kiwifruit planting duration sequence; S42, extracting the first kiwifruit planting time to the Qth kiwifruit planting time from the kiwifruit planting time sequence; S43. Extracting a first partition from the target area partition result, wherein the first partition has a first partition environmental parameter and a first partition growth parameter; S44, traversing the first kiwifruit planting time until the Qth kiwifruit planting time, combining them with the first partition environmental parameters, the first partition growth parameters, and the initial irrigation parameters, constructing growth constraints, performing representative sample statistics, obtaining representative sample growth parameters, and setting them as the first partition kiwifruit predicted growth sequence; S45. Add the predicted growth sequence of kiwifruit in the first partition to the kiwifruit growth prediction result.

[0052] In a preferred embodiment, first, the kiwifruit planting duration is obtained, and the future time zones are accumulated to obtain a kiwifruit planting duration sequence. The kiwifruit planting duration refers to the cumulative time length from the budding period of the kiwifruit to the current moment, usually in days or weeks. The current kiwifruit planting duration is accumulated with the time length of the future time zone to generate a continuous time series, i.e., the kiwifruit planting duration sequence. The kiwifruit planting duration sequence includes all time points from the current moment to the end of the future predicted time period, providing a time basis for subsequent growth prediction.

[0053] Next, the predicted time points are extracted from the kiwifruit planting time sequence. Specifically, the first kiwifruit planting time is extracted from the kiwifruit planting time sequence at preset time intervals, up to the Qth kiwifruit planting time, where Q is the total number of predicted time points. These extracted time points constitute the time nodes for growth prediction, and each time point corresponds to a specific kiwifruit growth stage.

[0054] Subsequently, the partition information to be predicted is extracted from the target area partition results. Taking the first partition as an example, the first partition is any one of the multiple partitions in the target area partition results. First, the first partition is selected as the prediction object. The first partition has first partition environmental parameters and first partition growth parameters. The first partition environmental parameters include environmental characteristic values ​​such as soil moisture, temperature, and light intensity for the partition. The first partition growth parameters include current growth monitoring values ​​of the kiwifruit in the partition, such as the number of leaves, internode length, and new shoot thickness, providing partition-specific basic data for growth prediction.

[0055] Next, the first kiwifruit planting time is traversed until the Qth kiwifruit planting time, and the growth constraints are constructed by combining them with the first partition environmental parameters, the first partition growth parameters, and the initial irrigation parameters. For each time point, the kiwifruit planting time corresponding to that time point is combined with the environmental parameters, growth parameters, and initial irrigation parameters of the first partition to form the growth constraint input for that time point. Subsequently, a representative sample statistical analysis of these growth constraints is performed using the target area representative sample growth prediction model to calculate the representative sample growth parameters of the corresponding time points. By traversing all time points and performing corresponding prediction calculations, a predicted growth sequence of kiwifruit in the first partition is obtained, which contains the growth prediction values ​​of the first partition at each future time point.

[0056] The predicted kiwifruit growth sequence for the first subarea is then added to the kiwifruit growth forecast results. This process is repeated for each subarea in the target area, gradually constructing a complete kiwifruit growth forecast. This kiwifruit growth forecast includes forecasts of future growth changes for each subarea, providing predictive data support for subsequent irrigation parameter optimization and distributed irrigation control.

[0057] Through time series-based growth prediction, the growth changes of kiwifruit in each zone in the future time period can be accurately predicted, laying a prediction foundation for achieving precise distributed irrigation control.

[0058] Furthermore, the first kiwifruit planting period is traversed until the Qth kiwifruit planting period, and the parameters are respectively combined with the first partition environmental parameters, the first partition growth parameters, and the initial irrigation parameters to construct growth constraints, perform representative sample statistics, and obtain representative sample growth parameters, which are set as the first partition kiwifruit predicted growth sequence, including: S441, collecting a first kiwifruit growth monitoring time series information set without variety difference; S442: based on the segmentation time step configured by the user terminal, traverse the first kiwifruit growth monitoring time series information set and randomly intercept to obtain multiple kiwifruit growth monitoring time series information; S443, traversing the plurality of kiwifruit growth monitoring time series information, extracting the kiwifruit growth monitoring information at the end time, setting it as supervision data, extracting the kiwifruit growth monitoring information at the start time, combining the planting environment record data, the planting duration at the end time, and the irrigation parameter record data as input data, and training the initial representative sample growth prediction model; S444. Collect a second kiwifruit growth monitoring time series information set of the kiwifruit variety in the target area, train the initial representative sample growth prediction model, obtain a target area representative sample growth prediction model, and embed it in the kiwifruit distributed irrigation controller.

[0059] In a preferred embodiment, first, a first kiwifruit growth monitoring time series information set without variety differences is collected. There is no variety difference problem in the first kiwifruit growth monitoring time series information set, that is, all kiwifruit growth monitoring data in the data set are from the same variety, thereby avoiding the interference of growth characteristics differences between different varieties on model training. The first kiwifruit growth monitoring time series information set is stored in the form of a time series, recording the changes in growth parameters such as the number of leaves, internode length, new shoot thickness, leaf shape index, etc. of this variety of kiwifruit at different growth stages, covering the complete growth cycle from the initial planting stage to the maturity stage. Each time series data point contains the growth parameter value at a specific time point and the corresponding time mark, forming a continuous time series data set. This data collection method without variety differences ensures that the initial representative sample growth prediction model can learn the pure growth change law without being affected by variety-specific factors, providing reliable basic data support for subsequent transfer learning training.

[0060] Then, based on the user-configured segmentation time step, the first kiwifruit growth monitoring time series information set is traversed and randomly intercepted to obtain multiple kiwifruit growth monitoring time series information. The segmentation time step is a time interval parameter pre-set by the user based on prediction accuracy requirements and computing resource constraints, typically ranging from several days to several weeks. The complete first kiwifruit growth monitoring time series information set is segmented according to this segmentation time step, with each segment containing a complete input-output data pair. This random interception process ensures the diversity and representativeness of the training data and avoids the temporal bias problem of time series data.

[0061] Subsequently, multiple kiwifruit growth monitoring time series were traversed to construct a training dataset. Specifically, for each time series segment, the kiwifruit growth monitoring information at the endpoint was extracted as supervision data, representing the target output to be predicted by the model. Simultaneously, the kiwifruit growth monitoring information at the starting point was extracted and combined with the corresponding planting environment records, planting duration, and irrigation parameter records at the endpoint to form a complete input data set, resulting in the training dataset. This input-output data pair construction method was used to train the initial representative sample growth prediction model using supervised learning methods.

[0062] Specifically, the initial representative sample growth prediction model utilizes a fully connected neural network architecture. The number of nodes in the network's input layer equals the total number of input parameters, including the sum of the dimensions of kiwifruit growth parameters (such as leaf number, internode length, and shoot diameter) at the starting point, recorded planting environment data (such as soil moisture, temperature, and light intensity), planting duration at the end point, and recorded irrigation parameters (such as irrigation frequency and intensity). The number of nodes in the network's output layer equals the dimensions of the kiwifruit growth parameters to be predicted, corresponding to the number of growth indicators at the end point, such as leaf number, internode length, and shoot diameter. The constructed training dataset is divided into training and validation sets according to a preset ratio, typically 8:2 or 7:3. The training process utilizes a batch gradient descent algorithm, with the training data fed into the neural network in batches for forward propagation. During the forward propagation process, the input data passes through each layer of the network, undergoing weight matrix operations and activation function processing to produce predicted growth parameter values ​​at the output layer. The mean squared error loss function (MSE) between the predicted value and the actual supervised data is calculated. This loss function reflects the model's current prediction error. Subsequently, a backpropagation algorithm is used to calculate the gradient of the loss function with respect to the weight parameters of each layer, and an optimization algorithm (such as the Adam optimizer or the SGD optimizer) is used to update the network weights. During the optimization process, the weight parameters are adjusted according to the set learning rate based on the calculated gradient information, gradually reducing the loss function value. This process of forward propagation-error calculation-backward propagation-weight update constitutes a complete training iteration cycle. Multiple training iterations are repeated, and after each cycle, the model performance is evaluated on the validation set to monitor the trend of the loss function and the prediction accuracy indicator. When the loss function on the validation set converges and the prediction accuracy meets the preset requirements, or when there is no significant improvement in verification performance over multiple consecutive cycles, the training process is terminated, resulting in a trained initial representative sample growth prediction model.

[0063] Subsequently, a second kiwifruit growth monitoring time-series information set for kiwifruit varieties in the target region was collected for transfer learning training of the initial representative sample growth prediction model. This second kiwifruit growth monitoring time-series information set specifically collects growth monitoring data for specific kiwifruit varieties within the target region. It is both region-specific and variety-specific, reflecting the environmental conditions of the target region and the growth characteristics of that specific variety. During the transfer learning process, the weight parameters of the trained fully connected layer in the initial representative sample growth prediction model remain unchanged, that is, the original fully connected layer structure is frozen to retain the general growth variation patterns learned from the first kiwifruit growth monitoring time-series information set. Specifically, transfer learning uses a progressive neuron expansion method for model adaptive training. First, a new neuron layer is added after the output layer of the initial representative sample growth prediction model. The input dimension of this new neuron layer is the same as the dimension of the output layer of the initial representative sample growth prediction model, and the output dimension remains consistent with the growth parameter dimension. During training, the second kiwifruit growth monitoring time-series information set is segmented according to the same segmentation time step to construct input-output data pairs for the specific varieties in the target region. The input data is first processed through a frozen fully connected layer to obtain an intermediate feature representation, which is then used as input for a new neural layer. This new neural layer corrects and optimizes the general prediction results of the initial representative sample growth prediction model by learning the growth variation characteristics of specific varieties in the target region. During training, a mean squared error loss function is used to calculate the prediction error between the output of the new neural layer and the actual growth data of the target region. The backpropagation algorithm is used to update only the weight parameters of the new neural layer, while the weights of the original fully connected layer remain unchanged. This training method preserves general growth prediction capabilities while incorporating the specific characteristics of the target region. After the training of the new neural layer converges, the prediction accuracy of the model is evaluated on the validation set. Subsequently, the performance of the current model on the growth prediction task for the target region is evaluated. If the prediction accuracy does not meet the preset requirements, the progressive expansion strategy is continued, adding a second neural layer after the current one. This second neural layer uses the output of the first one as input to continue learning deeper region- and variety-specific characteristics. Training is also performed by freezing the weights of the previous layer and training only the weights of the new layer. Repeat this progressive expansion and training process until the model's prediction accuracy on the validation set meets the preset requirements or the accuracy improvement stops being significant after multiple expansions. The trained model is then used as the representative sample growth prediction model for the target region. This model accurately predicts the growth changes of a specific kiwifruit variety in the target region under given environmental conditions, planting time, and irrigation parameters. This representative sample growth prediction model is then embedded in the kiwifruit distributed irrigation controller as the core computational module for subsequent growth prediction and irrigation parameter optimization.

[0064] Through transfer learning training with progressive neuron expansion, the general growth prediction rules are retained while the environmental specificity and variety specificity of the target area are incorporated, obtaining a high-precision growth prediction model suitable for the target area, providing reliable prediction support for achieving precise distributed irrigation control.

[0065] Furthermore, extracting the partitions where the kiwifruit growth prediction result is inconsistent with the expected kiwifruit growth, performing irrigation parameter optimization, obtaining distributed irrigation parameters, and performing kiwifruit distributed irrigation control, including: S511, extracting the predicted growth of kiwifruit in the first partition from the kiwifruit growth prediction result; S512: Compare the predicted growth of the kiwifruit in the first partition with the expected growth of the kiwifruit to obtain a kiwifruit growth deviation, wherein the kiwifruit growth deviation is equal to the Euclidean distance of the normalized values ​​of the kiwifruit growth parameters; S513. When the kiwifruit growth deviation is greater than or equal to the growth parameter deviation threshold, it is considered that the predicted growth of the kiwifruit in the first partition is inconsistent with the expected growth of the kiwifruit; otherwise, they are considered consistent.

[0066] In a preferred embodiment, first, the prediction data of a specific partition is extracted from the kiwifruit growth prediction result. Taking the first partition as an example, the first partition is any one of the multiple partitions in the target area partition result. The predicted growth of kiwifruit in the first partition is extracted from the kiwifruit growth prediction result. The predicted growth of kiwifruit in the first partition includes the predicted values ​​of growth parameters such as the number of leaves, internode length, new shoot thickness, leaf shape index, and fruit development status of the first partition in the future time period. These predicted values ​​are stored in a numerical form to provide basic data for subsequent growth deviation calculation.

[0067] Then, the predicted growth of kiwifruit in the first partition is compared with the expected growth of kiwifruit, and the degree of difference between the two is calculated. Specifically, the predicted growth of kiwifruit in the first partition and the expected growth of kiwifruit are normalized separately to eliminate the dimensional differences and numerical range differences between different growth parameters. The normalization process uses the minimum-maximum normalization method to map the values ​​of each growth parameter to the [0,1] interval. The calculation formula is: normalized value = (original value - minimum value) / (maximum value - minimum value). Among them, the minimum value and maximum value are the minimum monitored value and maximum monitored value of the growth parameter in the historical data, respectively. After normalization, the predicted growth of kiwifruit in the first partition and the expected growth of kiwifruit are converted into normalized vectors of the same dimension.

[0068] Subsequently, the Euclidean distance between the two normalized vectors is calculated to obtain the kiwifruit growth deviation. The kiwifruit growth deviation reflects the comprehensive difference between the predicted growth of kiwifruit in the first partition and the expected growth of kiwifruit. The larger the value, the more significant the difference. Then, the calculated kiwifruit growth deviation is compared with the preset growth parameter deviation threshold to determine the growth consistency. The growth parameter deviation threshold is a numerical parameter pre-set based on kiwifruit planting management requirements, growth control accuracy requirements and actual production experience. When the kiwifruit growth deviation is greater than or equal to the growth parameter deviation threshold, it indicates that there is a significant difference between the predicted growth of kiwifruit in the first partition and the expected growth of kiwifruit, and the predicted growth of kiwifruit in the first partition is considered to be inconsistent with the expected growth of kiwifruit. At this time, the partition needs to optimize the irrigation parameters and improve the growth prediction results by adjusting the irrigation strategy to make it closer to the expected growth. When the kiwifruit growth deviation is less than the growth parameter deviation threshold, the difference between the predicted and expected kiwifruit growth in the first subarea is within an acceptable range, and the predicted and expected kiwifruit growth in the first subarea are considered consistent. At this point, the current irrigation parameter settings can be maintained for that subarea, without the need for additional parameter optimization.

[0069] Through the calculation of growth deviation and threshold comparison based on Euclidean distance, the partitions where the growth prediction results are inconsistent with the expected growth can be accurately identified, providing a clear optimization target for the subsequent optimization of irrigation parameters and ensuring the accuracy and effectiveness of distributed irrigation control.

[0070] Furthermore, extracting the partitions where the kiwifruit growth prediction result is inconsistent with the expected kiwifruit growth, performing irrigation parameter optimization, obtaining distributed irrigation parameters, and performing kiwifruit distributed irrigation control, including: S521, extracting the first partition where the kiwifruit growth prediction result is inconsistent with the kiwifruit expected growth; S522, initializing irrigation parameters to obtain several irrigation parameters; S523, traversing the plurality of irrigation parameters, performing growth prediction on the first subarea, and obtaining a plurality of predicted growth conditions of the first subareas; S524, traversing the predicted growth conditions of the plurality of first partitions, comparing the predicted growth conditions with the expected growth conditions of the kiwifruit, and obtaining growth deviations of the kiwifruit in the plurality of first partitions; S525. When the growth deviations of the kiwifruit in the first subareas are all greater than or equal to the growth parameter deviation threshold, extracting, from the plurality of irrigation parameters, irrigation parameters with minimum growth deviation, irrigation parameters with median growth deviation, and irrigation parameters with maximum growth deviation based on the growth deviations of the kiwifruit in the first subareas, wherein the irrigation parameters with median growth deviation are grouped into one or two groups, and the irrigation parameters with minimum growth deviation and the irrigation parameters with maximum growth deviation are grouped into one group. S526: construct a first search axis with the minimum growth deviation irrigation parameter as the target point and the median growth deviation irrigation parameter as the starting point; S527: construct a second search axis with the irrigation parameter with the minimum growth deviation as the target point and the irrigation parameter with the maximum growth deviation as the starting point; S528: constructing a third search axis with the median growth deviation irrigation parameter as the target point and the maximum growth deviation irrigation parameter as the starting point; S529: Perform a Z-shaped search along the first search axis, the second search axis, and the third search axis respectively to obtain updated irrigation parameters, and execute a loop.

[0071] In a preferred embodiment, all subregions requiring irrigation parameter optimization are first screened from the kiwifruit growth prediction results, and irrigation parameter updates and optimization are performed on these subregions. One subregion is sequentially selected from all subregions requiring irrigation parameter optimization and processed as the first subregion. Here, the first subregion represents any subregion requiring irrigation parameter optimization. Then, the irrigation parameters for the first subregion are initialized, generating several sets of candidate irrigation parameters for subsequent optimization, resulting in a number of irrigation parameters. To improve the efficiency and accuracy of irrigation parameter optimization, the irrigation parameter initialization process utilizes a constraint generation method based on healthy kiwifruit samples of the same variety. Specifically, irrigation parameter records for healthy kiwifruit samples of the same kiwifruit variety and in good growth condition as those in the target area are first obtained through historical data retrieval and sample screening. These healthy samples should meet criteria such as excellent growth indicators, freedom from pests and diseases, and stable yields, ensuring that their irrigation parameters are valuable for reference. Then, based on the current kiwifruit growth period in the first subregion, irrigation parameter records for the same growth period are extracted from a healthy sample database, including historical values ​​for key parameters such as irrigation frequency, duration, intensity, and water temperature. Next, a central tendency analysis was performed on the irrigation parameters of the selected healthy samples during the same growth period. Statistical methods were used to calculate the statistical characteristics of each irrigation parameter, such as the mean, standard deviation, and quantile, and to analyze the distribution pattern and central tendency of the parameter values. Based on the statistical analysis results, a reasonable distribution range for each irrigation parameter was determined, typically using the mean ± 2 times the standard deviation or the 10th to 90th percentile as the parameter distribution range. This parameter distribution range not only reflects the irrigation parameter characteristics of the healthy samples but also avoids interference from extreme values, thereby limiting the optimization search space for the irrigation parameters. Based on the determined parameter distribution range, a random sampling method was used to generate several irrigation parameters within the constraint range of each parameter. Each set of irrigation parameters contained a complete set of irrigation control parameter settings, and all parameter values ​​were within the constraint range of the healthy samples, ensuring the rationality and feasibility of the initial parameters. This constrained initialization method can effectively improve the convergence speed and optimization effect of irrigation parameter optimization.

[0072] Subsequently, the growth prediction evaluation of the generated irrigation parameters is performed one by one. Specifically, each set of irrigation parameters is selected in turn, and it is combined with the environmental parameters, current growth parameters and weather forecast data of the first partition, and the growth prediction calculation is performed using the representative sample growth prediction model of the target area. During the prediction process, the initial irrigation parameters of the first partition are replaced by the current irrigation parameters, and other input conditions are kept unchanged. The growth changes of kiwifruit in the first partition in the future time period under the irrigation parameters are calculated using the representative sample growth prediction model of the target area. Each set of irrigation parameters corresponds to an independent growth prediction result, which includes the predicted values ​​of key growth indicators such as the number of leaves, internode length, and new shoot thickness. By traversing all initialized irrigation parameters, the predicted growth of several first partitions is obtained, which forms the evaluation basis for subsequent parameter optimization.

[0073] Next, the growth prediction effect corresponding to each set of irrigation parameters was evaluated. Specifically, the predicted growth of each first partition was selected in turn and quantitatively compared with the expected growth of kiwifruit. Using the Euclidean distance calculation method, the growth parameters were normalized and the deviation between the predicted growth and the expected growth was calculated to obtain the growth deviation of kiwifruit in the first partition. This reflects the degree of closeness between the growth prediction result under the current irrigation parameter settings and the expected target. The smaller the growth deviation of kiwifruit in the first partition, the better the irrigation effect of the current irrigation parameters. By traversing all the predicted growth of the first partition, several growth deviations of kiwifruit in the first partition were obtained, providing a quantitative basis for subsequent parameter screening and search direction determination.

[0074] When the growth deviations corresponding to all irrigation parameters fail to achieve the expected results, further parameter optimization search is required. Specifically, when the growth deviations of several kiwifruit in the first partition are greater than or equal to the growth parameter deviation threshold, it indicates that the currently initialized irrigation parameters cannot meet the growth requirements and a more in-depth parameter optimization search is required. At this point, based on the order of the growth deviations of the kiwifruit in the first partition, key reference points are screened from the corresponding irrigation parameters: the irrigation parameter with the minimum growth deviation, the irrigation parameter with the median growth deviation, and the irrigation parameter with the maximum growth deviation. Specifically, first, all irrigation parameters are sorted in ascending order of growth deviation value, and the irrigation parameter with the minimum growth deviation is extracted as the irrigation parameter with the minimum growth deviation, which represents the relatively optimal solution under the current conditions. Then, the irrigation parameter with the maximum growth deviation is extracted as the irrigation parameter with the maximum growth deviation, which represents the worst solution under the current conditions. Next, the irrigation parameter with the median growth deviation is determined. When the total number of irrigation parameters is an odd number, the median corresponds to a unique middle position, and the irrigation parameters with the median growth deviation are grouped together. When the total number of irrigation parameters is even, the median corresponds to the two middle irrigation parameters. In this case, the median deviation irrigation parameters for growth are divided into two groups. These median deviation irrigation parameters represent the middle level of the current parameter distribution and provide a balanced reference point for subsequent searches. Through parameter screening based on deviation sorting, the irrigation parameters with minimum growth deviation, median growth deviation, and maximum growth deviation were determined, laying the foundation for developing subsequent parameter search strategies.

[0075] Subsequently, a multi-directional search axis is constructed based on the screened key reference points to provide directional guidance for the subsequent parameter space search. When constructing the first search axis, the irrigation parameter with the minimum growth deviation is used as the target point, and the irrigation parameter with the median growth deviation is used as the starting point. The direction vector of the first search axis points to the parameter space direction from the medium level to the current optimal solution, representing the improvement path from the average effect to the better effect. When constructing the second search axis, the irrigation parameter with the minimum growth deviation is used as the target point, and the irrigation parameter with the maximum growth deviation is used as the starting point. The direction vector of the second search axis points to the parameter space direction from the worst solution to the current optimal solution, representing a significant improvement path from poor effect to good effect. When constructing the third search axis, the irrigation parameter with the median growth deviation is used as the target point, and the irrigation parameter with the maximum growth deviation is used as the starting point. The direction vector of the third search axis points to the parameter space direction from the worst solution to the medium level, representing a gradual improvement path from poor effect to average effect. Each search axis defines a specific search direction in the multidimensional irrigation parameter space. These axes cover multiple possible improvement directions starting from the current known parameter points, providing a search strategy for comprehensive parameter space exploration.

[0076] A Z-shaped search strategy is then executed along the three constructed search axes (i.e., the first, second, and third search axes) to achieve efficient exploration and optimization of the parameter space. Z-shaped search is a non-strict axis search method. Compared to a strictly linear search along the axes, it can cover a larger search range and increase the probability of finding the global optimal solution. Specifically, the Z-shaped search employs a zigzag search path along each axis. Taking the first search axis as an example, the search process begins at the starting point of the growth median deviation irrigation parameter, advances along the axis by a certain step length, then deviates a certain distance perpendicular to the axis, and then continues along the axis, repeating this process to form a Z-shaped search trajectory. This search method maintains the overall trend along the axis while expanding the exploration of the neighborhood space around the axis. During the search process, each search position corresponds to a new set of irrigation parameters. For each new set of irrigation parameters, growth prediction is performed using the growth prediction model for representative samples of the target area, and the corresponding growth deviation value is calculated. The optimal irrigation parameters found during the search are recorded, i.e., the parameter combination with the minimum growth deviation. The same Z-shaped search process is performed on the second search axis and the third search axis at the same time. The parallel search of the three axes ensures full coverage of the parameter space and avoids the local optimal problem caused by a single search direction. All new parameter combinations obtained during the search process constitute the updated irrigation parameters. The updated irrigation parameters are used as the input for a new round of optimization, and the loop process from S523 to S529 is repeated. In each loop, the search axis and search strategy are re-determined based on the current optimal parameter distribution, and the global optimal irrigation parameter combination is gradually approached. When the improvement in the growth deviation after multiple consecutive cycles is less than the preset convergence threshold, or the search reaches the preset maximum number of iterations, the loop process is terminated. The optimal irrigation parameters obtained at this time are the distributed irrigation parameters of the first partition, which can make the growth prediction results of the first partition closest to the expected growth target.

[0077] Through the parameter optimization method based on multi-axis Z-shaped search, the optimal irrigation parameter combination can be efficiently found in the high-dimensional parameter space, providing reliable parameter optimization support for achieving precise distributed irrigation control.

[0078] Example 2, as Figure 2 As shown, based on the same inventive concept as the kiwifruit smart irrigation method using distributed optimization control provided in Example 1, an embodiment of the present invention further provides a kiwifruit smart irrigation system using distributed optimization control, comprising: A data acquisition module 11 is used to obtain weather forecast data for the target area in the future time zone, initial irrigation parameters and expected growth of kiwifruit; The dual partitioning module 12 is used to perform environmental partitioning on the target area to obtain a first partitioning result, and to perform kiwifruit growth partitioning on the target area to obtain a second partitioning result; A partition intersection module 13 is configured to intersect the first partition result and the second partition result to obtain a target area partition result; A growth prediction module 14 is configured to traverse the target area partition results, extract the kiwifruit monitoring growth in the partition, and perform growth prediction in combination with the weather forecast data and the initial irrigation parameters to obtain a kiwifruit growth prediction result, wherein the kiwifruit growth prediction result corresponds one-to-one to the target area partition results; The optimization control module 15 is used to extract the partitions where the kiwifruit growth prediction results are inconsistent with the expected kiwifruit growth, perform irrigation parameter optimization, obtain distributed irrigation parameters, and perform kiwifruit distributed irrigation control.

[0079] Furthermore, the dual partition module 12 executes the following steps: Collecting the first environmental attribute monitoring value distribution information of the target area, performing cluster analysis, and obtaining the first environmental attribute partitioning result; Until the Nth environmental attribute monitoring value distribution information of the target area is collected, cluster analysis is performed, and the Nth environmental attribute partition result is obtained; Intersecting the first environmental attribute partition results up to the Nth environmental attribute partition result to obtain the first partition result.

[0080] Furthermore, the dual partition module 12 may further execute the following steps: Extracting a first position environmental attribute monitoring value and a second position environmental attribute monitoring value from the first environmental attribute monitoring value distribution information, wherein the first position and the second position are adjacent positions; Calculating an environmental attribute monitoring value deviation between the environmental attribute monitoring value at the first location and the environmental attribute monitoring value at the second location; When the environmental attribute monitoring value deviation is less than a first environmental attribute deviation threshold, the first position and the second position are merged into a third position. At the same time, the average of the environmental attribute monitoring value of the first position and the environmental attribute monitoring value of the second position is calculated and set as the environmental attribute monitoring value of the third position. The first environmental attribute deviation threshold is a user-preset value. When the environmental attribute monitoring value deviation is greater than or equal to a first environmental attribute deviation threshold, the first position and the second position are regarded as different positions; When the deviations of the environmental attribute monitoring values ​​of any two adjacent positions are greater than or equal to the first environmental attribute deviation threshold, the first environmental attribute partitioning result is output.

[0081] Furthermore, the execution steps of the growth prediction module 14 include: Obtaining the kiwifruit planting time, accumulating the future time zone, and obtaining a kiwifruit planting time sequence; Extract the first kiwifruit planting time to the Qth kiwifruit planting time from the kiwifruit planting time sequence; Extracting a first partition from the target area partition result, wherein the first partition has a first partition environmental parameter and a first partition growth parameter; Traversing the first kiwifruit planting time until the Qth kiwifruit planting time, respectively combining them with the first partition environmental parameters, the first partition growth parameters, and the initial irrigation parameters, constructing growth constraints, performing representative sample statistics, and obtaining representative sample growth parameters, which are set as the first partition kiwifruit predicted growth sequence; The predicted growth sequence of kiwifruit in the first partition is added to the kiwifruit growth prediction result.

[0082] Furthermore, the execution steps of the growth prediction module 14 also include: Collect the first kiwifruit growth monitoring time series information set without cultivar differences; Based on the segmentation time step configured by the user end, traverse the first kiwifruit growth monitoring time series information set and randomly intercept to obtain multiple kiwifruit growth monitoring time series information; Traversing the plurality of kiwifruit growth monitoring time series information, extracting the kiwifruit growth monitoring information at the end time, setting it as supervision data, extracting the kiwifruit growth monitoring information at the start time, combining the planting environment record data, the planting time at the end time, and the irrigation parameter record data as input data, and training the initial representative sample growth prediction model; A second kiwifruit growth monitoring time series information set of the kiwifruit variety in the target area is collected, the initial representative sample growth prediction model is trained, and a target area representative sample growth prediction model is obtained, which is embedded in the kiwifruit distributed irrigation controller.

[0083] Furthermore, the execution steps of the optimization control module 15 include: Extracting the predicted growth of kiwifruit in the first partition from the kiwifruit growth prediction result; Comparing the predicted growth of the kiwifruit in the first partition with the expected growth of the kiwifruit to obtain a kiwifruit growth deviation, wherein the kiwifruit growth deviation is equal to the Euclidean distance of the normalized values ​​of the kiwifruit growth parameters; When the kiwifruit growth deviation is greater than or equal to the growth parameter deviation threshold, it is considered that the predicted growth of the kiwifruit in the first partition is inconsistent with the expected growth of the kiwifruit; otherwise, they are considered consistent.

[0084] Furthermore, the execution steps of the optimization control module 15 also include: Extracting the first partition where the kiwifruit growth prediction result is inconsistent with the kiwifruit expected growth; Initialize the irrigation parameters and obtain several irrigation parameters; Traversing the plurality of irrigation parameters, performing growth prediction on the first subarea, and obtaining a plurality of predicted growth conditions of the first subareas; Traversing the predicted growth conditions of the plurality of first partitions, comparing the predicted growth conditions with the expected growth conditions of the kiwifruit, and obtaining growth deviations of the kiwifruit in the plurality of first partitions; When the growth deviations of the kiwifruits in the first partitions are all greater than or equal to the growth parameter deviation threshold, based on the growth deviations of the kiwifruits in the first partitions, extracting the minimum growth deviation irrigation parameter, the median growth deviation irrigation parameter and the maximum growth deviation irrigation parameter from the plurality of irrigation parameters, wherein the median growth deviation irrigation parameter is one or two groups, and the minimum growth deviation irrigation parameter and the maximum growth deviation irrigation parameter are both one group; Taking the minimum growth deviation irrigation parameter as the target point and the median growth deviation irrigation parameter as the starting point, a first search axis is constructed; Taking the irrigation parameter with the minimum growth deviation as the target point and the irrigation parameter with the maximum growth deviation as the starting point, a second search axis is constructed; Taking the median growth deviation irrigation parameter as the target point and the maximum growth deviation irrigation parameter as the starting point, a third search axis is constructed; A Z-shaped search is performed along the first search axis, the second search axis, and the third search axis to obtain updated irrigation parameters, and a loop is executed.

[0085] It should be noted that, in the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0086] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0087] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0088] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0089] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0090] Although preferred embodiments of the present invention have been described, additional changes and modifications to these embodiments may occur to those skilled in the art once the basic inventive concepts become known.

[0091] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the present invention and its equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A kiwifruit smart irrigation method using distributed optimization control, characterized in that: Applied to kiwifruit distributed irrigation controller, including: Obtain weather forecast data for the target area in the future time zone, initial irrigation parameters, and expected kiwifruit growth conditions; Perform environmental zoning on the target area to obtain the first zoning result, and perform kiwifruit growth zoning on the target area to obtain the second zoning result; Intersecting the first partitioning result and the second partitioning result to obtain a target area partitioning result; Traversing the target area partition results, extracting the monitored growth of kiwifruit in the partition, and performing growth prediction in combination with the weather forecast data and the initial irrigation parameters to obtain a kiwifruit growth prediction result, wherein the kiwifruit growth prediction result corresponds one-to-one to the target area partition results; Extract the partitions where the kiwifruit growth prediction result is inconsistent with the expected kiwifruit growth, perform irrigation parameter optimization, obtain distributed irrigation parameters, and perform kiwifruit distributed irrigation control.

2. The method according to claim 1, wherein Perform environmental zoning on the target area and obtain the first zoning result, including: Collecting the first environmental attribute monitoring value distribution information of the target area, performing cluster analysis, and obtaining the first environmental attribute partitioning result; Until the Nth environmental attribute monitoring value distribution information of the target area is collected, cluster analysis is performed, and the Nth environmental attribute partition result is obtained; Intersecting the first environmental attribute partition results up to the Nth environmental attribute partition result to obtain the first partition result.

3. The method according to claim 2, wherein Collecting the first environmental attribute monitoring value distribution information of the target area, performing cluster analysis, and obtaining the first environmental attribute partitioning result, including: Extracting a first position environmental attribute monitoring value and a second position environmental attribute monitoring value from the first environmental attribute monitoring value distribution information, wherein the first position and the second position are adjacent positions; Calculating an environmental attribute monitoring value deviation between the environmental attribute monitoring value at the first location and the environmental attribute monitoring value at the second location; When the environmental attribute monitoring value deviation is less than a first environmental attribute deviation threshold, the first position and the second position are merged into a third position. At the same time, the average of the environmental attribute monitoring value of the first position and the environmental attribute monitoring value of the second position is calculated and set as the environmental attribute monitoring value of the third position. The first environmental attribute deviation threshold is a user-preset value. When the environmental attribute monitoring value deviation is greater than or equal to a first environmental attribute deviation threshold, the first position and the second position are regarded as different positions; When the deviations of the environmental attribute monitoring values ​​of any two adjacent positions are greater than or equal to the first environmental attribute deviation threshold, the first environmental attribute partitioning result is output.

4. The method according to claim 1, wherein Traversing the target area partition results, extracting the kiwifruit monitoring growth in the partition, combining the weather forecast data and the initial irrigation parameters, performing growth forecasting, and obtaining kiwifruit growth forecast results, including: Obtaining the kiwifruit planting time, accumulating the future time zone, and obtaining a kiwifruit planting time sequence; Extract the first kiwifruit planting time to the Qth kiwifruit planting time from the kiwifruit planting time sequence; Extracting a first partition from the target area partition result, wherein the first partition has a first partition environmental parameter and a first partition growth parameter; Traversing the first kiwifruit planting time until the Qth kiwifruit planting time, respectively combining them with the first partition environmental parameters, the first partition growth parameters, and the initial irrigation parameters, constructing growth constraints, performing representative sample statistics, and obtaining representative sample growth parameters, which are set as the first partition kiwifruit predicted growth sequence; The predicted growth sequence of kiwifruit in the first partition is added to the kiwifruit growth prediction result.

5. The method according to claim 4, wherein Traversing the first kiwifruit planting time until the Qth kiwifruit planting time, respectively combining them with the first partition environmental parameters, the first partition growth parameters, and the initial irrigation parameters, constructing growth constraints, performing representative sample statistics, and obtaining representative sample growth parameters, which are set as the first partition kiwifruit predicted growth sequence, including: Collect the first kiwifruit growth monitoring time series information set without cultivar differences; Based on the segmentation time step configured by the user end, traverse the first kiwifruit growth monitoring time series information set and randomly intercept to obtain multiple kiwifruit growth monitoring time series information; Traversing the plurality of kiwifruit growth monitoring time series information, extracting the kiwifruit growth monitoring information at the end time, setting it as supervision data, extracting the kiwifruit growth monitoring information at the start time, combining the planting environment record data, the planting time at the end time, and the irrigation parameter record data as input data, and training the initial representative sample growth prediction model; A second kiwifruit growth monitoring time series information set of the kiwifruit variety in the target area is collected, the initial representative sample growth prediction model is trained, and a target area representative sample growth prediction model is obtained, which is embedded in the kiwifruit distributed irrigation controller.

6. The method according to claim 1, wherein Extracting partitions where the kiwifruit growth prediction result is inconsistent with the expected kiwifruit growth, performing irrigation parameter optimization, obtaining distributed irrigation parameters, and performing kiwifruit distributed irrigation control, including: Extracting the predicted growth of kiwifruit in the first partition from the kiwifruit growth prediction result; Comparing the predicted growth of the kiwifruit in the first partition with the expected growth of the kiwifruit to obtain a kiwifruit growth deviation, wherein the kiwifruit growth deviation is equal to the Euclidean distance of the normalized values ​​of the kiwifruit growth parameters; When the kiwifruit growth deviation is greater than or equal to the growth parameter deviation threshold, it is considered that the predicted growth of the kiwifruit in the first partition is inconsistent with the expected growth of the kiwifruit; otherwise, they are considered consistent.

7. The method according to claim 1, wherein Extracting partitions where the kiwifruit growth prediction result is inconsistent with the expected kiwifruit growth, performing irrigation parameter optimization, obtaining distributed irrigation parameters, and performing kiwifruit distributed irrigation control, including: Extracting the first partition where the kiwifruit growth prediction result is inconsistent with the kiwifruit expected growth; Initialize the irrigation parameters and obtain several irrigation parameters; Traversing the plurality of irrigation parameters, performing growth prediction on the first subarea, and obtaining a plurality of predicted growth of the first subareas; Traversing the predicted growth conditions of the plurality of first partitions, comparing the predicted growth conditions with the expected growth conditions of the kiwifruit, and obtaining growth deviations of the kiwifruit in the plurality of first partitions; When the growth deviations of the kiwifruits in the first partitions are all greater than or equal to the growth parameter deviation threshold, based on the growth deviations of the kiwifruits in the first partitions, extracting the minimum growth deviation irrigation parameter, the median growth deviation irrigation parameter and the maximum growth deviation irrigation parameter from the plurality of irrigation parameters, wherein the median growth deviation irrigation parameter is one or two groups, and the minimum growth deviation irrigation parameter and the maximum growth deviation irrigation parameter are both one group; Taking the minimum growth deviation irrigation parameter as the target point and the median growth deviation irrigation parameter as the starting point, a first search axis is constructed; Taking the irrigation parameter with the minimum growth deviation as the target point and the irrigation parameter with the maximum growth deviation as the starting point, a second search axis is constructed; Taking the median growth deviation irrigation parameter as the target point and the maximum growth deviation irrigation parameter as the starting point, a third search axis is constructed; A Z-shaped search is performed along the first search axis, the second search axis, and the third search axis to obtain updated irrigation parameters, and a loop is executed.

8. A kiwifruit smart irrigation system using distributed optimization control, characterized in that: A method for implementing the method according to any one of claims 1 to 7, applied to a kiwifruit distributed irrigation controller, comprising: A data acquisition module is used to obtain weather forecast data for the target area in the future time zone, initial irrigation parameters and expected growth of kiwifruit; The dual partitioning module is used to perform environmental partitioning on the target area to obtain the first partitioning result, and to perform kiwifruit growth partitioning on the target area to obtain the second partitioning result; A partition intersection module, configured to intersect the first partition result and the second partition result to obtain a target area partition result; A growth prediction module is used to traverse the target area partition results, extract the kiwifruit monitoring growth in the partition, combine the weather forecast data and the initial irrigation parameters, perform growth prediction, and obtain a kiwifruit growth prediction result, wherein the kiwifruit growth prediction result corresponds one-to-one with the target area partition results; The optimization control module is used to extract the partitions where the kiwifruit growth prediction results are inconsistent with the expected kiwifruit growth, perform irrigation parameter optimization, obtain distributed irrigation parameters, and perform kiwifruit distributed irrigation control.

Images