An intelligent fruit tree cultivation method and system based on the Internet of Things

Abstract

The application discloses a kind of fruit tree intelligent cultivation method and system based on Internet of Things, specifically related to fruit tree cultivation Internet of Things field, including reading each collection position in garden in current management cycle and last management cycle Internet of Things monitoring data, and aligning according to the fruit tree position corresponding to each collection position, generate position record, according to the position record of each collection position, calculate the root area change value, heat and humidity change value and growth change value of each collection position in current management cycle relative to last management cycle, generate change record;The application aligns the calculation, segment division and fragmented execution of the cross-cycle change of different collection positions in the garden, to solve the problem that the difference of fruit trees in different positions under the fixed partition unified control in the prior art cannot be directly converted into corresponding management action.

Description

Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) technology in fruit tree cultivation, and more specifically, to an IoT-based intelligent fruit tree cultivation method and system. Background Technology

[0002] In fruit tree cultivation and seed and seedling management, existing technologies typically involve setting up data collection points in greenhouses or orchards to collect soil temperature and humidity, nutrients, air temperature and humidity, light, and tree growth status. The collected data is then summarized by greenhouse area, plot, or irrigation branch, and control instructions such as drip irrigation and fertilization, ventilation and shading, and pest and disease control are issued based on this data to improve the timeliness of water and fertilizer regulation and growth management. Under the parallel management of semi-enclosed greenhouses and open-field orchards in Shanghai, the automated system must maintain continuous data collection and control while directly adapting to existing irrigation branches and environmental control units to complete the execution. However, there are often significant differences in root zone moisture changes, local heat and humidity conditions, and tree growth responses near vents, side films, drip irrigation ends, low-lying wet areas, and canopy closure areas. In this case, if the data is still aggregated and uniformly controlled according to fixed zones, it is easy to encounter situations where some tree positions have excessive root zone water content after continuous watering, leading to increased risks of waterlogging and disease, while adjacent tree positions still have insufficient watering and slow growth recovery. On-site, this manifests as inconsistent response results at different locations within the same control unit. The reason for this problem is that the data collection stage has already reflected the actual differences between different locations, while the control stage still uses fixed greenhouse areas, fixed plots, or fixed branches as the execution objects, making it impossible to directly correspond local differences to local management actions. Therefore, the technical problem to be solved by this application is: how to identify the actual differences of fruit trees in different locations within the same orchard without changing the existing irrigation and environmental control zones during the Internet of Things intelligent cultivation of fruit trees, and to directly translate these differences into management actions for the corresponding locations, so as to avoid excessive or insufficient local management caused by unified control. Summary of the Invention

[0003] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide an intelligent fruit tree cultivation method and system based on the Internet of Things, which solves the problems mentioned in the background art by performing alignment calculations, segment division, and piecewise execution on cross-cycle changes at different collection locations within the orchard.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a smart fruit tree cultivation method based on the Internet of Things, comprising: S1. Read the IoT monitoring data of each collection location in the park during the current management cycle and the previous management cycle, and align them according to the fruit tree locations corresponding to each collection location to generate location records; S2. Based on the location records of each collection point, calculate the root zone change value, heat and humidity change value, and growth change value of each collection point in the current management cycle relative to the previous management cycle, and generate change records. S3. Within the same irrigation zone and environmental control zone, compare the change records in the adjacent order of each collection location, and merge consecutive collection locations with the same root zone change direction, heat and humidity change direction and growth change direction into location segments to generate a location segment sequence. S4. Based on the change records corresponding to each location segment, determine the cultivation actions for each location segment. When the root zone change value decreases, the heat and humidity change value increases, and the growth change value decreases, it is determined to be a water and fertilizer irrigation action. When the root zone change value does not decrease, the heat and humidity change value increases, and the growth change value decreases, it is determined to be a green control action. When the root zone change value does not decrease, the heat and humidity change value decreases, and the growth change value increases, it is determined to be a shaping and pruning action. Generate execution instructions. S5. Send the execution instructions corresponding to each location segment to the corresponding irrigation branch and environmental control zone for execution, and update the location segment sequence and execution instructions according to the IoT monitoring data after execution, and output the cultivation results for the next management cycle.

[0005] In a preferred embodiment, S1 includes: S1-1. Read the IoT monitoring data of the current management cycle and the previous management cycle according to the collection location, and arrange them in ascending order according to the collection time to generate a collection sequence; S1-2. For the collection sequence at each collection location, match the IoT monitoring data of the current management cycle with the IoT monitoring data of the previous management cycle according to the corresponding category of IoT monitoring data, and write the unmatched IoT monitoring data into the missing marker to generate cycle alignment results. S1-3. Associate the period alignment results of each collection location with the corresponding fruit tree location to generate a location record.

[0006] In a preferred embodiment, S2 includes: S2-1. Read the location records of each collection location, extract the root zone data, heat and humidity data and growth data corresponding to the same collection location in the current management cycle and the previous management cycle, and generate a periodic data group. S2-2. For the periodic data groups at each collection location, sort the current management cycle data and the previous management cycle data in ascending order of collection time, and pair them one by one in the order of arrangement to generate root zone pairing group, heat and humidity pairing group and growth pairing group. S2-3. For the root zone pairing group, heat and humidity pairing group and growth pairing group at each collection location, calculate the difference item by subtracting the data of the previous management cycle from the data of the current management cycle. Then, calculate the root zone change value, heat and humidity change value and growth change value by dividing the sum of the differences by the number of pairing items.

[0007] In a preferred embodiment, S2 further includes: S2-4. Determine the root region change value that is greater than zero as the root region change direction increasing, the root region change value that is less than zero as the root region change direction decreasing, and the root region change value that is equal to zero as the root region change direction remaining unchanged. Values ​​of heat and humidity change greater than zero are defined as indicating an increasing direction of heat and humidity change; values ​​of heat and humidity change less than zero are defined as indicating a decreasing direction of heat and humidity change; and values ​​of heat and humidity change equal to zero are defined as indicating an unchanged direction of heat and humidity change. Growth change values ​​greater than zero are defined as increasing growth direction, growth change values ​​less than zero are defined as decreasing growth direction, and growth change values ​​equal to zero are defined as unchanged growth direction. The change record is composed of root zone change value, heat and humidity change value, growth change value, root zone change direction, heat and humidity change direction, and growth change direction.

[0008] In a preferred embodiment, S3 includes: S3-1. Read the change records of each collection location within the same irrigation zone and environmental control zone, form adjacent location pairs according to the adjacent order of each collection location, and compare the root zone change direction, heat and humidity change direction, and growth change direction of each adjacent location pair respectively. Write a continuous mark when all three are consistent, write a check mark when only two of the three are consistent, and write a boundary mark in other cases to generate the initial result. S3-2. For each adjacent position pair corresponding to the mark to be inspected, read the initial results of the previous adjacent position pair and the next adjacent position pair respectively. If both the previous and next adjacent position pairs are written with continuous marks, and the current inconsistency item is in the same direction in both the previous and next adjacent position pairs, rewrite the current mark to be inspected as a continuous mark. Otherwise, rewrite it as a boundary mark and generate the verification result.

[0009] In a preferred embodiment, S3 further includes: S3-3. Read the initial results and verification results in the order of each collection location. If there are both continuous markers and boundary markers at the same adjacent location, retain the boundary markers. Otherwise, retain the markers in the verification results. If only the initial results exist, retain the initial results and generate the final marking results. S3-4. Read the final marking results in the order of each collection position, merge the collection positions connected end to end by consecutive markings into candidate segments, and take the next collection position corresponding to the boundary marking as the starting position of the next candidate segment to generate a candidate segment sequence. S3-5. For two adjacent candidate segments in the candidate segment sequence, compare the change records of the last acquisition position of the previous candidate segment and the first acquisition position of the next candidate segment. If the change direction of the root region, the change direction of heat and humidity, and the change direction of growth are all consistent, they are merged into the same position segment; otherwise, they are kept separate to generate a position segment sequence.

[0010] In a preferred embodiment, S4 includes: S4-1. Read the change records corresponding to each location segment, and extract the root zone change value, heat and humidity change value and growth change value of each collection location according to the location segment to generate segment change group; S4-2. For each segment of the fragment, determine the segment change group item by item. The sampling location where the root zone change value decreases, the heat and humidity change value increases, and the growth change value decreases is determined as the corresponding water and fertilizer irrigation action. The sampling location where the root zone change value does not decrease, the heat and humidity change value increases, and the growth change value decreases is determined as the corresponding green prevention and control action. The sampling location where the root zone change value does not decrease, the heat and humidity change value decreases, and the growth change value increases is determined as the corresponding shaping and pruning action. Generate the results corresponding to the sampling location.

[0011] In a preferred embodiment, S4 further includes: S4-3. For each location segment, count the number of collection locations corresponding to water and fertilizer irrigation actions, green pest control actions, and shaping and pruning actions in the results. If the number of collection locations is greater than the number of collection locations of the other two, the action corresponding to the number of collection locations is determined as the cultivation action of that location segment. If the number of collection locations of two or three is equal, count the sum of the absolute values ​​of root zone change, heat and humidity change, and growth change for the corresponding collection locations. If the sum of one of these values ​​is greater than the sum of the other two, the action corresponding to this sum is determined as the cultivation action of that location segment. If the sums are equal, determine the cultivation action of that location segment in the order of water and fertilizer irrigation actions, green pest control actions, and shaping and pruning actions, and generate the action results. S4-4. Associate the action results of each location segment with the corresponding location segment, irrigation branch and environmental control zone to generate execution instructions.

[0012] In a preferred embodiment, S5 includes: S5-1: Read the execution instructions corresponding to each location segment, and send them to the corresponding irrigation branch and environmental control zone for execution according to the location segment, and generate an execution record; S5-2. Read the IoT monitoring data corresponding to the execution record after execution, align it according to the fruit tree position corresponding to each collection location, regenerate the position record, change record and position segment sequence, and generate the updated result; S5-3. Based on the update results, redetermine the execution instructions corresponding to each position segment, and combine the updated position segment sequence and execution instructions to form the cultivation results for the next management cycle.

[0013] In a preferred embodiment, an Internet of Things-based intelligent fruit tree cultivation system includes: The alignment generation module is used to read IoT monitoring data from each collection location in the park during the current management cycle and the previous management cycle, and align them according to the corresponding fruit tree locations to generate location records. The change calculation module calculates the root zone change value, heat and humidity change value, and growth change value of each collection location relative to the previous management cycle in the current management cycle, based on the location records of each collection location, and generates change records. The segmentation module is used to compare change records in the adjacent order of each collection location within the same irrigation zone and environmental control zone. It merges consecutive collection locations with the same root zone change direction, heat and humidity change direction, and growth change direction into location segments, generating a location segment sequence. The action determination module determines the cultivation action for each location segment based on the change records corresponding to each location segment. Specifically, when the root zone change value decreases, the heat and humidity change value increases, and the growth change value decreases, it is determined to be a water and fertilizer irrigation action. When the root zone change value does not decrease, the heat and humidity change value increases, and the growth change value decreases, it is determined to be a green control action. When the root zone change value does not decrease, the heat and humidity change value decreases, and the growth change value increases, it is determined to be a shaping and pruning action, and an execution instruction is generated. The feedback update module is used to send the execution instructions corresponding to each location segment to the corresponding irrigation branch and environmental control zone for execution, and update the location segment sequence and execution instructions according to the IoT monitoring data after execution, and output the cultivation results for the next management cycle.

[0014] The technical effects and advantages of this invention are as follows: 1. By calculating change records based on the collection location and outputting management actions based on location segments, the actual differences in different locations within the park can be directly converted into corresponding control results, thereby relatively alleviating the problem of excessive or insufficient local management caused by the unified execution of fixed zones; 2. By organizing the IoT monitoring data of the current management cycle and the previous management cycle according to the collection location, collection time and data category, a consistent input can be provided for subsequent calculation of change values, thereby relatively improving the impact of mixed cross-cycle values ​​on the judgment results. 3. By comparing the root zone change direction, heat and humidity change direction, and growth change direction of adjacent sampling locations, and merging continuous and consistent locations into location segments, it is possible to identify local continuous areas within existing irrigation zones and environmental control zones, thereby relatively improving the correspondence between management actions and on-site conditions. 4. By determining irrigation, green pest control, or pruning actions based on changes within location segments, and combining this with statistical results from the locations collected within the segments, the actions can be determined, which can relatively reduce the interference of fluctuations in a single location on the overall action selection. 5. By sending execution commands to the corresponding irrigation branches and environmental control zones for execution, and regenerating location records, change records, and location segment sequences after execution, a continuous write-back management link can be formed, thereby improving the continuity of actions in the next management cycle. Attached Figure Description

[0015] Figure 1 This is a flowchart of the method steps of the present invention.

[0016] Figure 2 This is a schematic diagram of the system modules of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Refer to the attached diagram in the instruction manual. Figures 1-2 The present invention provides an intelligent fruit tree cultivation method based on the Internet of Things, comprising: S1. Read the IoT monitoring data of each collection location in the park during the current management cycle and the previous management cycle, and align them according to the fruit tree locations corresponding to each collection location to generate location records; In this implementation, IoT monitoring data generated at the same collection location within two management cycles are first organized into directly comparable data. The organized data is then mapped to the fruit tree location and used as input for subsequent calculations of root zone change values, heat and humidity change values, and growth change values. During processing, IoT monitoring data from the current management cycle and the previous management cycle are first read according to the collection location. Then, the data sequence of the same collection location is established according to the collection time. Based on this, the correspondence between the two management cycles is completed according to the data category. Finally, the corresponding results are associated with the corresponding fruit tree location. The implementation process includes the following steps: First, IoT monitoring data from the current and previous management cycles are read separately according to the collection location, and then sorted in ascending order by collection time to generate a collection sequence. In specific processing, a unique location identifier is first assigned to each collection location within the park. Then, using the location identifier as the search condition, all IoT monitoring data generated at that collection location within the current and previous management cycles are extracted from the edge collection terminal or monitoring database. Each IoT monitoring data entry includes at least the collection location identifier, collection time, data category, and data value. The current management cycle is the cycle corresponding to the current cultivation action to be performed, and the previous management cycle is the cycle immediately preceding the current management cycle. After reading, data with missing collection location identifiers or missing collection times are first removed. Then, data from the same collection location within the current management cycle are sorted by collection time from earliest to latest, and data from the same collection location within the previous management cycle are also sorted by collection time from earliest to latest, forming the current management cycle collection sequence and the previous management cycle collection sequence for that collection location. If two IoT monitoring data entries with the same data category exist at the same collection time, the one written last is retained. Next, for the collection sequences at each collection location, the IoT monitoring data of the current management cycle is paired with the IoT monitoring data of the previous management cycle according to the corresponding category of IoT monitoring data. Unpaired IoT monitoring data is written into a missing tag to generate cycle alignment results. In specific processing, the collection sequence is first split according to data category within each collection location. Data of the same category is assigned to the corresponding category sequence. Data used for root zone change value calculation is assigned to the root zone category, data used for heat and humidity change value calculation is assigned to the heat and humidity category, and data used for growth change value calculation is assigned to the growth category. After the category splitting is completed, the current management cycle category sequence and the previous management cycle category sequence under the same collection location and the same data category are paired item by item in the order of arrangement, the first to the first, the second to the second, until one category sequence ends. Data items that have been matched are written into the pairing results, and the remaining data for which no matching item is found are written into a missing tag. The missing tag includes at least the collection location identifier, management cycle identifier, data category, and missing status. After the above processing, each collection location generates a cycle alignment result containing paired data items and missing tags. Subsequently, the periodic alignment results of each collection location are associated with the corresponding fruit tree location to generate a location record. Specifically, a pre-established table mapping collection locations to fruit tree locations within the park is first called, where each collection location corresponds to one fruit tree location. When multiple collection locations correspond to the same fruit tree location, the periodic alignment results of each collection location are retained, and the same fruit tree location identifier is written during the association process. During the association process, the periodic alignment result of each collection location and its corresponding fruit tree location identifier are written into the same record, and the collection location identifier, current management cycle identifier, previous management cycle identifier, and matching results and missing markers for each category are simultaneously retained to form a location record. After this processing, each location record can simultaneously correspond to a specific collection location and a fruit tree location. Subsequent steps can directly obtain the data results that have been completed between the same collection location, the same fruit tree location, and two management cycles when reading the location record. Through the above processing, sorting can be completed first within the collection location, then the correspondence between two management cycles and the same category of data can be completed within the same collection location, and finally the correspondence can be mapped to the fruit tree location, so that subsequent calculations can be directly based on the already sorted location records, avoiding mixing data of different categories, different locations or different cycles together for value taking. In practical applications: Taking a semi-enclosed greenhouse blueberry growing area in Shanghai as an example, each soil and tree sampling node near a drip irrigation point can be considered as a sampling location. The current management cycle is set as the current irrigation control cycle, and the previous management cycle is set as the previous irrigation control cycle. The system first reads the soil moisture, air temperature and humidity, light, and shoot growth data of the sampling location within two management cycles and arranges them in the order of sampling time. Then, it matches the data of the same category in the two management cycles in sequence. If shoot growth data is missing at a certain time in the previous management cycle, a missing mark is written under that category. Then, the cycle alignment result of the sampling location is associated with the corresponding blueberry plant location to form a location record for subsequent change value calculation. For the case where one fruit tree location corresponds to two sampling locations in an open-field vineyard, the location records of the two sampling locations are retained respectively, and the two location records are associated with the same grapevine location simultaneously, so that subsequent change calculation and segmentation can be completed without confusing the sampling locations.

[0019] S2. Based on the location records of each collection point, calculate the root zone change value, heat and humidity change value, and growth change value of each collection point in the current management cycle relative to the previous management cycle, and generate change records. In this embodiment, the purpose of processing S2 is to further transform the data from the two management cycles in the location record into a change record that can be directly used for segment division and cultivation action determination. Specifically, the root zone data, heat and humidity data, and growth data under the current management cycle and the previous management cycle are extracted from the location record of each collection location. Then, they are reordered according to the collection time within their respective categories and paired item by item. After that, the difference between each item is calculated by subtracting the data from the previous management cycle from the data of the current management cycle, and the average value of each difference is used as the change value of the corresponding category. Finally, the change direction is determined according to the positive and negative relationship of the change value, and the change value and the change direction are written into the same change record. After this processing, subsequent steps can directly read the change record to complete the consistency comparison between adjacent collection locations. The implementation process includes the following steps: First, read the location records of each collection point, and extract the root zone data, heat and humidity data, and growth data corresponding to the same collection point in the current and previous management cycles to generate a cycle data group. In specific processing, for each collection point, read the collection point identifier, fruit tree location identifier, current management cycle identifier, previous management cycle identifier, category pairing results, and missing data markers from its location record. Within the same collection point, split the data into root zone data, heat and humidity data, and growth data according to data category. Root zone data consists of data in the location record that has been assigned to the root zone category and whose corresponding cycle has been completed; heat and humidity data consists of data in the location record that has been assigned to the heat and humidity category and whose cycle has been completed. The data corresponding to the completed cycle are the growth data, which are the data in the location record that have been classified into the growth category and whose cycle has been completed. For data items with missing markers, they are not included in the cycle data group of the current collection location. When all items of a certain category are missing markers at a certain collection location, no corresponding data group is generated for that category, and the missing status is retained in the location record of that collection location so that subsequent steps can continue to calculate according to the existing categories and not mistakenly merge missing items into normal values. After the extraction is completed, each collection location forms three types of cycle data groups: root zone, heat and humidity, and growth. Each cycle data group contains two parts: the data of the current management cycle and the data of the previous management cycle. Next, for the periodic data groups at each acquisition location, the current management cycle data and the previous management cycle data are arranged in ascending order of acquisition time, and then paired item by item according to the arrangement order to generate root zone pairing groups, heat and humidity pairing groups, and growth pairing groups. Specifically, within each acquisition location, the current management cycle data in the root zone data group is arranged from earliest to latest acquisition time, and the previous management cycle data is arranged from earliest to latest acquisition time, then paired item by item according to the arrangement order, one for each pair, until one group is completed. The heat and humidity data group and the growth data group use the same... Processing method: If the number of data items in the current management cycle is greater than the number of data items in the previous management cycle, the data exceeding the current management cycle will not be included in the pairing group; if the number of data items in the previous management cycle is greater than the number of data items in the current management cycle, the data exceeding the previous management cycle will also not be included in the pairing group; the number of pairing items is determined by the smaller value of the number of data items in the two groups; after the above processing, each collection location will obtain a root zone pairing group, a heat and humidity pairing group, and a growth pairing group, and each item in each pairing group consists of data from the same collection location, the same category, and in the same arrangement order within the two management cycles; Subsequently, for each sampling location, the root zone pairing group, heat and humidity pairing group, and growth pairing group are calculated item by item by subtracting the data from the previous management cycle from the current management cycle data. The root zone change value, heat and humidity change value, and growth change value are then calculated by dividing the sum of these differences by the number of pairing items. Specifically, within each sampling location, for each pairing item in the root zone pairing group, the current management cycle data value is subtracted from the previous management cycle data value to obtain the root zone difference. All root zone differences are then summed, and the sum is divided by the number of pairing items in the root zone pairing group to obtain the root zone change value for that sampling location. The heat and humidity pairing group and the growth pairing group are calculated using the same method. If the number of pairing items in a pairing group is zero, no change value is generated for that category at that sampling location, and this category remains missing in subsequent change records without being replaced by a zero value. Through this process, each sampling location can obtain root zone change values, heat and humidity change values, and growth change values, thereby transforming similar data from two management cycles into directly comparable change results. Then, the following parameters are defined: Root zone change values ​​greater than zero are defined as increasing direction of root zone change; root zone change values ​​less than zero are defined as decreasing direction of root zone change; and root zone change values ​​equal to zero are defined as unchanged direction of root zone change. Similarly, heat and humidity change values ​​greater than zero are defined as increasing direction of heat and humidity change; heat and humidity change values ​​less than zero are defined as decreasing direction of heat and humidity change; and heat and humidity change values ​​equal to zero are defined as unchanged direction of heat and humidity change. Furthermore, growth change values ​​greater than zero are defined as increasing direction of growth change; growth change values ​​less than zero are defined as decreasing direction of growth change; and growth change values ​​equal to zero are defined as unchanged direction of growth change. These parameters are then combined to form a variable... The process involves recording data; specifically, within each sampling location, the relationship between the root zone change value and zero is determined first, and then the direction of root zone change is output; the relationship between the heat and humidity change value and zero is determined in the same way, and the direction of heat and humidity change is output; the relationship between the growth change value and zero is determined, and the direction of growth change is output; if no change value is generated for a certain category, the change direction for that category is not output, and the category remains missing in the change record; after the direction is determined, the sampling location identifier, fruit tree location identifier, root zone change value, heat and humidity change value, growth change value, root zone change direction, heat and humidity change direction, and growth change direction are written into the same record to form the change record corresponding to that sampling location, which is used for subsequent comparison and segmentation by adjacent sampling locations; Through the above processing, the cross-cycle corresponding data in the location record can be transformed into change values ​​and change directions, and each collection location can form an independent, complete and retrievable change record, thereby providing a unified input for subsequent segmentation and cultivation action determination; In practical applications: Taking a semi-enclosed greenhouse blueberry growing area in Shanghai as an example, a certain sampling location generates three sets of soil moisture data in the current management cycle and three sets of soil moisture data in the previous management cycle. The system first pairs the data item by item according to the order of sampling time, and then calculates the difference between the three items. If the sum of the three differences divided by three results in a negative value, the root zone change value at that sampling location is recorded as the corresponding value, and the direction of root zone change is determined to be decreasing. The air temperature and humidity data and shoot growth data at the same sampling location are also paired and calculated in the same way. If the temperature and humidity change value is positive and the growth change value is negative, the direction of temperature and humidity change is determined to be increasing and the direction of growth change is decreasing, respectively. Finally, the three types of change values ​​and the three types of change directions are written into the change record of that sampling location. For cases where a certain sampling location in an open-field vineyard lacks a growth data item in the previous management cycle, only the successfully paired data item is used to calculate the growth change value. Unpaired items are not included in the calculation to avoid directly incorporating missing data into the calculation result.

[0020] S3. Within the same irrigation zone and environmental control zone, compare the change records in the adjacent order of each collection location, and merge consecutive collection locations with the same root zone change direction, heat and humidity change direction and growth change direction into location segments to generate a location segment sequence. In this embodiment, the purpose of processing S3 is to convert the change records of each collection location within the same irrigation zone and environmental control zone into a location segment sequence, so that the determination of subsequent cultivation actions is based on the continuous changes of adjacent locations. Specifically, the adjacent order of each collection location within the same irrigation zone and environmental control zone is first determined, and adjacent location pairs are formed accordingly. Then, the root zone change direction, heat and humidity change direction, and growth change direction of each adjacent location pair are compared to form an initial result, and then the part to be inspected is verified. Subsequently, the initial result and the verification result are merged into the final labeling result, and candidate segments are segmented according to the final labeling result. Finally, a supplementary comparison is performed on the candidate segments that are adjacent to each other to obtain the location segment sequence. The implementation process includes the following steps: First, the change records of each sampling location within the same irrigation zone and environmental control zone are read. Adjacent location pairs are formed according to the adjacent order of the sampling locations, and the root zone change direction, heat and humidity change direction, and growth change direction of each adjacent location pair are compared. A continuous marker is written when all three are consistent; a check marker is written when only two of the three are consistent; and a boundary marker is written in other cases, generating the initial results. In specific processing, the sampling location identifiers and change records of all sampling locations within the same irrigation zone and environmental control zone are extracted first. Then, the sampling locations are arranged according to a pre-determined adjacent order. The adjacent order is determined by the arrangement of the sampling locations on the corresponding irrigation branch or environmental control path. The first sampling location and the second sampling location form a pair. The first adjacent position pair, the second acquisition position, and the third acquisition position form the second adjacent position pair, and so on, generating all adjacent position pairs in sequence. For each adjacent position pair, the root zone change direction, heat and humidity change direction, and growth change direction of the two acquisition positions are read respectively. When all three directions are consistent, a continuous marker is written to the adjacent position pair. When only two of the three directions are consistent and the other is inconsistent, a check marker is written. When only one of the three directions is consistent or all three are inconsistent, a boundary marker is written. After comparing all adjacent position pairs, the initial result is generated. If there is only one acquisition position in a certain irrigation zone and environmental control zone, no adjacent position pair is formed, and the acquisition position is directly used as the starting position of a separate segment for subsequent processing. Furthermore, for each adjacent position pair corresponding to the mark to be inspected, the initial results of the previous adjacent position pair and the next adjacent position pair are read respectively. If both the previous and next adjacent position pairs are written with continuous marks, and the current inconsistency item remains in the same direction in both the previous and next adjacent position pairs, the current mark to be inspected is rewritten as a continuous mark; otherwise, it is rewritten as a boundary mark, and a review result is generated. In specific processing, the adjacent position pairs corresponding to all marks to be inspected are first located in the initial results, and then the previous and next adjacent position pairs are read for each adjacent position pair. For the mark to be inspected at the beginning, since there is no previous adjacent position pair, it is directly rewritten as a boundary mark. For the mark to be inspected at the end, since there is no next adjacent position pair, it is also directly rewritten as a boundary mark. For the mark to be inspected that has both a previous and next adjacent position pair, the previous adjacent position pair is first determined. If both the current and next adjacent position pairs are written with continuous markers in the initial result, then the current marker to be inspected is rewritten as a boundary marker. If so, then it is determined whether the inconsistency item in the current marker to be inspected belongs to the root region change direction, the heat and humidity change direction, or the growth change direction. Then, the change direction of the corresponding item in the previous and next adjacent position pairs is read respectively, and it is determined whether the change direction is in the same direction as the same side of the sampling position in the current marker to be inspected. Here, "in the same direction" means that the change direction of the corresponding item has not changed direction between the three adjacent position pairs, such as all increasing, all decreasing, or all remaining unchanged. If the inconsistency item is in the same direction in both the previous and next adjacent position pairs, it means that the current inconsistency item belongs to a local transition, and the current marker to be inspected is rewritten as a continuous marker. Otherwise, the current marker to be inspected is rewritten as a boundary marker. After all the markers to be inspected are processed, a review result is generated. Subsequently, the initial and verification results are read crosswise according to the order of each collection location. When both continuous and boundary markers exist in the same adjacent location pair, the boundary marker is retained; otherwise, the marker in the verification result is retained. When only the initial result exists, the initial result is retained, generating the final marking result. In specific processing, the initial results are read one by one according to the order of adjacent location pairs, and it is checked whether there is a verification result for the adjacent location pair. If there is no corresponding record for the adjacent location pair in the verification result, it means that the adjacent location pair is not the marker to be checked in the initial result, and the marker in the initial result is directly retained. If there is a corresponding record for the adjacent location pair in the verification result, the marker in the initial result and the marker in the verification result are merged for judgment. When both continuous and boundary markers exist in the same adjacent location pair, the boundary marker is retained. When there is a verification result and the above situation does not occur, the marker in the verification result is retained. After processing each pair, only one final marker is retained for each adjacent location pair, and all final markers constitute the final marking result. After this processing, the direct continuity relationship, direct boundary relationship in the initial result, and the result after verification are all unified into the same result set. Next, the final marking results are read in the order of each collection location. Collection locations connected end-to-end by consecutive markings are merged into candidate segments. The collection location following the boundary marker is taken as the starting position of the next candidate segment, generating a candidate segment sequence. In specific processing, the first collection location after sorting is taken as the starting position of the first candidate segment, and then the final marking results are read sequentially according to adjacent position pairs. When the current adjacent position pair is written with a consecutive marker, the next collection location is merged into the current candidate segment. When the current adjacent position pair is written with a boundary marker, the current candidate segment ends, and the collection location following the boundary marker is taken as the starting position of the next candidate segment. After processing all adjacent position pairs in sequence, a candidate segment sequence arranged in the order of collection locations is obtained. If a candidate segment contains only one collection location, it is retained as a single-location candidate segment and is not deleted. This ensures that all collection locations within the same irrigation zone and environmental control zone are included in the candidate segment sequence, and each collection location appears only once in the candidate segment sequence. Finally, for two adjacent candidate segments in the candidate segment sequence, the change records of the last acquisition position of the preceding candidate segment and the first acquisition position of the following candidate segment are compared. If the root region change direction, heat and humidity change direction, and growth change direction are all consistent, they are merged into the same position segment; otherwise, they are kept separate, generating a position segment sequence. In specific processing, the preceding and following candidate segments are read sequentially according to the candidate segment sequence. For the connection position between the two, that is, the last acquisition position of the preceding candidate segment and the first acquisition position of the following candidate segment, their change records are read again and the root region change direction, heat and humidity change direction, and growth change direction are compared. If all three directions are consistent, the preceding and following candidate segments are merged into the same position segment; if any one of the directions is inconsistent, they are kept separate. After merging, new adjacent candidate segments are processed until the candidate segment sequence processing is completed, and finally, a position segment sequence is generated. This step is used to eliminate local segmentation formed in the review process, so that truly continuous acquisition positions are included in the same position segment, while acquisition positions with directional boundaries are kept separate. Through the above processing, the direction of change of adjacent collection locations can be compared for the first time, then the transition locations that are consistent can be checked, then the results can be uniformly marked and the candidate segments can be segmented, and finally a first-to-last connection judgment can be added. In this way, the collection locations with continuous direction of change within the same irrigation zone and environmental control zone can be merged into a location segment sequence. The resulting location segment sequence retains the boundary position and avoids directly cutting off the local transition position, and can be directly used for subsequent cultivation action determination steps. In practical applications: Taking a drip irrigation branch in a semi-enclosed greenhouse blueberry growing area in Shanghai as an example, if there are six sampling locations within the same irrigation zone and environmental control zone, arranged from the first to the sixth sampling location according to the branch direction, then five adjacent location pairs are first formed; among them, the first, second, and fourth adjacent location pairs have the same three change directions, the third adjacent location pair only has different heat and humidity change directions, and the fifth adjacent location pair has only one consistent item among the three items, then continuous markers, inspection markers, and boundary markers are written respectively; for the third adjacent location pair, if both its preceding and following adjacent location pairs are written with continuous markers, and the heat and humidity change direction is in the same direction... If the direction remains the same during the comparison, it is rewritten as a continuous marker; otherwise, it is rewritten as a boundary marker. After all verifications, the system merges the first to fourth collection positions into a single candidate segment according to the final marking result, divides the fifth and sixth collection positions into subsequent candidate segments, and then performs a supplementary comparison on the connection positions of the candidate segments to finally obtain the position segment sequence used for determining cultivation actions. For cases where the undulating terrain in open-field vineyards causes a short-term shift in the direction of heat and humidity changes at intermediate collection positions, the above verification steps can distinguish the local shift that occurs only at one position from the true boundary, avoiding the misclassification of the same continuous management area into multiple segments.

[0021] S4. Based on the change records corresponding to each location segment, determine the cultivation actions for each location segment. When the root zone change value decreases, the heat and humidity change value increases, and the growth change value decreases, it is determined to be a water and fertilizer irrigation action. When the root zone change value does not decrease, the heat and humidity change value increases, and the growth change value decreases, it is determined to be a green control action. When the root zone change value does not decrease, the heat and humidity change value decreases, and the growth change value increases, it is determined to be a shaping and pruning action. Generate execution instructions. In this implementation, the purpose of S4 is to uniformly determine the cultivation action that should be performed in the current management cycle for each location segment based on the change records of each collection location within the location segment, and to implement the cultivation action into a sendable execution command. Specifically, the root zone change value, heat and humidity change value, and growth change value of all collection locations within the location segment are extracted first, and then each collection location is determined to correspond to a certain type of cultivation action. Based on this, the number of collection locations corresponding to each type of action is counted for each location segment. If the number of collection locations corresponding to a certain type of action is dominant, it is directly determined as the cultivation action of that location segment. If there are cases where the number is equal, the sum of the absolute values ​​of the root zone change value, the absolute values ​​of the heat and humidity change value, and the absolute values ​​of the growth change value of the corresponding collection location is compared. If the sum is still equal, the determination is completed in a predetermined order. Finally, the action result is associated with the corresponding location segment, irrigation branch, and environmental control zone. The implementation process includes the following steps: First, read the change records corresponding to each location segment. Extract the root zone change value, heat and humidity change value, and growth change value for each collection location according to the location segment, generating a segment change group. Specifically, first, read the collection location identifier contained in each location segment sequentially according to the location segment sequence. Then, retrieve the root zone change value, heat and humidity change value, and growth change value corresponding to that collection location identifier from the change record, and write them into the same data group according to the order of the collection location within that location segment, forming the segment change group corresponding to that location segment. Each data entry in the segment change group must contain at least the collection location identifier, root zone change value, heat and humidity change value, and growth change value. If a collection location lacks any change value in the change record, that collection location will not enter the current segment change group, and the missing state will be retained at the corresponding location. If all collection locations in a location segment lack change values, that location segment will not enter the cultivation action determination process of the current management cycle, but will maintain its original state in subsequent execution instructions. Through the above processing, each location segment forms a segment change group composed of valid collection locations within the segment. Subsequently, the segment change groups at each location were evaluated item by item. Locations where the root zone change value decreased, the heat and humidity change value increased, and the growth change value decreased were identified as corresponding irrigation / fertilization actions. Locations where the root zone change value remained unchanged, the heat and humidity change value increased, and the growth change value decreased were identified as corresponding green pest control actions. Locations where the root zone change value remained unchanged, the heat and humidity change value decreased, and the growth change value increased were identified as corresponding shaping / pruning actions. Results corresponding to each location were generated. In specific processing, the combination relationship of three types of change values ​​was sequentially determined for each location in the segment change group: when the root zone change value was less than zero, the heat and humidity change value was greater than zero, and the growth change value was less than zero, the location was considered a suitable location. The sampling location is set as the corresponding water and fertilizer irrigation action; when the root zone change value is greater than or equal to zero, the heat and humidity change value is greater than zero, and the growth change value is less than zero, the sampling location is set as the corresponding green control action; when the root zone change value is greater than or equal to zero, the heat and humidity change value is less than zero, and the growth change value is greater than zero, the sampling location is set as the corresponding shaping and pruning action; for sampling locations that do not match the above three types of combinations, the corresponding sampling location result is not written and is not included in the subsequent action statistics; after completing the item-by-item judgment, each location segment generates a corresponding sampling location result, which specifies which sampling locations in the location segment correspond to water and fertilizer irrigation action, green control action, and shaping and pruning action respectively; Next, for each location segment, the number of collection locations corresponding to irrigation, green pest control, and pruning actions is counted. If the number of collection locations is greater than the number of collection locations in the other two locations, the action corresponding to that collection location is identified as the cultivation action for that location segment. If the number of collection locations in two or three locations is equal, the sum of the absolute values ​​of root zone change, heat and humidity change, and growth change is counted for each location. If the sum of one of these sums is greater than the sum of the other two, the action corresponding to that sum is identified as the cultivation action for that location segment. If the sums are equal, the cultivation action for that location segment is determined in the order of irrigation, green pest control, and pruning, generating the action results. In specific processing, three numbers are counted for each location segment: the number of collection locations corresponding to irrigation, green pest control, and pruning actions. When there are two other numbers, the action corresponding to that number is directly written as the cultivation action for that location segment. If two or three numbers are equal, the sum is calculated only for the collection locations corresponding to the parallel actions. When calculating the sum, the absolute values ​​of root zone change, heat and humidity change, and growth change are calculated for each collection location corresponding to a certain parallel action. Then, the sums of the above values ​​for all collection locations corresponding to the parallel action are added together to obtain the total value corresponding to that action. If one of the total values ​​is greater than the other two, the action corresponding to that total value is written as the cultivation action for that location segment. If the total values ​​are still equal, the actions are compared in the order of water and fertilizer irrigation, green control, and shaping and pruning, with the former being the priority, and written as the cultivation action for that location segment. After the above processing is completed, each location segment generates a unique action result. The order here is used to eliminate the final conflict in the parallel state, with water and fertilizer irrigation in the first place, green control in the second place, and shaping and pruning in the third place. Finally, the action results of each location segment are associated with the corresponding location segment, irrigation branch, and environmental control zone to generate execution instructions. Specifically, the segment identifier of each location segment in the location segment sequence is read first, then the correspondence between the collection location and the irrigation branch and environmental control zone is called, summarizing the irrigation branches and environmental control zones to which all collection locations in the same location segment belong. When all collection locations in the same location segment correspond to the same irrigation branch and the same environmental control zone, the action results are directly associated with that irrigation branch and that environmental control zone. When multiple irrigation branches or multiple environmental control zones are involved in the same location segment, they are split and written to the corresponding execution objects according to the attribution relationship of the collection location in that location segment. After association, the location segment identifier, cultivation action, irrigation branch identifier, environmental control zone identifier, and current management cycle identifier are written into the same execution data to form an execution instruction. The execution instructions generated in this way can be directly used for subsequent sending and execution steps without needing to re-determine the action type and execution object of the location segment. Through the above processing, the change value can be extracted within the location segment first, and then the action attribution can be determined according to the collection location. Then, the unique cultivation action can be determined by comparing the number, the sum, and the order. Finally, the cultivation action is written into the corresponding execution object to form an execution instruction that can be directly issued. This avoids the problem of directly averaging the changes of different collection locations within the segment before making a judgment, and also avoids the problem of multiple actions falling on the same location segment at the same time and being unable to be executed. In practical applications: Taking a semi-enclosed greenhouse blueberry growing area in Shanghai as an example, a certain location segment contains three sampling points. Two of these sampling points satisfy the following conditions: root zone change value is less than zero, heat and humidity change value is greater than zero, and growth change value is less than zero. The third sampling point satisfies the following conditions: root zone change value is greater than or equal to zero, heat and humidity change value is greater than zero, and growth change value is less than zero. Therefore, in this location segment, the number of sampling points corresponding to water and fertilizer irrigation is two, the number corresponding to green pest control is one, and the number corresponding to pruning is zero. The system directly identifies the cultivation action in this location segment as water and fertilizer irrigation and assigns this action... The system associates the drip irrigation branch and the greenhouse environment control zone corresponding to the location segment with the execution command. If two sampling locations within a certain location segment in an open vineyard correspond to greening and pruning actions respectively, and the two numbers are equal, the system continues to calculate the sum of the absolute values ​​of the three types of changes at the sampling locations corresponding to these two actions. The sum is used to determine which type of action the location segment will ultimately execute. If the sums are still equal, the former of the greening and pruning actions or the former of the pruning actions will be used as the cultivation action for the location segment in a predetermined order, thereby ensuring that each location segment outputs only one executable command in the current management cycle.

[0022] S5. Send the execution instructions corresponding to each location segment to the corresponding irrigation branch and environmental control zone for execution, and update the location segment sequence and execution instructions according to the IoT monitoring data after execution, and output the cultivation results for the next management cycle; In this embodiment, the purpose of S5 is to implement the execution instructions corresponding to the determined location segments to the corresponding irrigation branches and environmental control zones, and after the execution is completed, to reread the IoT monitoring data and update it according to the aforementioned generation rules of location records, change records and location segment sequences, thereby obtaining cultivation results that can be directly called in the next management cycle. In specific processing, the execution instructions are first sent according to the location segments and the execution process information is retained. Then, the location records, change records and location segment sequences are regenerated according to the IoT monitoring data after execution. Finally, the execution instructions corresponding to each location segment are re-determined based on the updated results. The implementation process includes the following steps: First, the execution instructions corresponding to each location segment are read and sent to the corresponding irrigation branches and environmental control zones for execution, generating execution records. Specifically, the execution instructions for each location segment are read sequentially according to the location segment sequence. Each execution instruction includes at least the location segment identifier, cultivation action, irrigation branch identifier, environmental control zone identifier, and current management cycle identifier. Then, execution instructions for cultivation actions involving water and fertilizer irrigation are sent to the corresponding irrigation branches, and those for cultivation actions involving green pest control or pruning are sent to the corresponding environmental control zones. When the same location segment corresponds to multiple irrigation branches or multiple environmental control zones, the instructions are sent to the corresponding execution objects according to the affiliation of each collection location within that location segment. After sending, the sending time, receiving object, execution start time, and execution completion time of each execution instruction are recorded to form an execution record. If an execution object does not return execution completion information, an incomplete status is written to the execution record, and subsequent update processing of the location segment corresponding to that execution object is stopped. Next, the IoT monitoring data corresponding to the execution record is read, aligned with the fruit tree positions corresponding to each collection location, and the location record, change record, and location segment sequence are regenerated to generate an update result. Specifically, the execution completion time corresponding to the execution completion status is extracted from the execution record, and the IoT monitoring data formed between the execution completion time and the end time of the current management cycle is used as the IoT monitoring data after execution. For multiple IoT monitoring data generated at the same collection location within this time period, they are still arranged in ascending order of collection time. After reading, according to the processing rules in S1, the IoT monitoring data after execution is compared with its immediately preceding data. The IoT monitoring data for one management cycle is aligned, and location records are regenerated. Then, according to the processing rules in S2, the root zone change value, heat and humidity change value, growth change value, and corresponding change direction are calculated based on the regenerated location records, and change records are regenerated. Then, according to the processing rules in S3, the change directions between adjacent collection locations are compared based on the regenerated change records, and location segment sequences are regenerated. The original collection location identifier, fruit tree location identifier, irrigation zone identifier, and environmental control zone identifier remain unchanged throughout the above regeneration process. After all processing is completed, the regenerated location records, change records, and location segment sequences are combined to form the update result. Subsequently, the execution instructions corresponding to each location segment are redefined based on the update results, and the updated location segment sequence and execution instructions are combined to form the cultivation results for the next management cycle. Specifically, the updated location segment sequence and its corresponding change records are first read from the update results. Then, according to the processing rules in S4, the cultivation action for each updated location segment is redefined, and the corresponding execution instructions are generated. When an updated location segment no longer contains a valid acquisition location in the update results, no new execution instructions are generated. When the cultivation action corresponding to an updated location segment is consistent with the cultivation action of the previous management cycle, a new execution instruction is still generated, and its management cycle identifier is rewritten to the identifier of the next management cycle. After all updates are completed, the updated location segment sequence and execution instructions are written into the same result set according to the location segments, forming the cultivation results for the next management cycle, which can be directly called for the next round of reading and execution. Through the above processing, the execution instructions corresponding to the current location segment can be implemented to the specific execution object first, and then the IoT monitoring data after execution can be reintegrated into the generation process of location record, change record and location segment sequence. Finally, the cultivation results that can be used in the next management cycle are formed, so that the entire cultivation process can be continuously connected between management cycles. In practical applications: Taking a semi-enclosed greenhouse blueberry growing area in Shanghai as an example, after a certain location segment is determined to be subject to water and fertilizer irrigation in the current management cycle, the system sends the execution command corresponding to that location segment to the corresponding drip irrigation branch and records the sending time, drip irrigation start time, and drip irrigation end time. After the drip irrigation ends and before the end of the current management cycle, the system rereads the soil moisture, air temperature and humidity, and shoot growth data collected at each location segment, and regenerates the location record, change record, and location segment sequence according to the aforementioned rules. Then, it re-determines whether the location segment will continue to perform water and fertilizer irrigation, or change to green pest control or pruning in the next management cycle. For a certain location segment in an open-field vineyard that is performing green pest control in the current management cycle, the execution command is sent to the corresponding environmental control zone. After execution, the system also reads the IoT monitoring data after execution, regenerates the location record, change record, and location segment sequence, and writes the new location segment sequence and execution command into the cultivation results of the next management cycle, so that the next management cycle can continue to operate directly based on the updated state.

[0023] Furthermore, an IoT-based intelligent fruit tree cultivation system includes: The alignment generation module is used to read IoT monitoring data from each collection location in the park during the current management cycle and the previous management cycle, and align them according to the corresponding fruit tree locations to generate location records. The change calculation module calculates the root zone change value, heat and humidity change value, and growth change value of each collection location relative to the previous management cycle in the current management cycle, based on the location records of each collection location, and generates change records. The segmentation module is used to compare change records in the adjacent order of each collection location within the same irrigation zone and environmental control zone. It merges consecutive collection locations with the same root zone change direction, heat and humidity change direction, and growth change direction into location segments, generating a location segment sequence. The action determination module determines the cultivation action for each location segment based on the change records corresponding to each location segment. Specifically, when the root zone change value decreases, the heat and humidity change value increases, and the growth change value decreases, it is determined to be a water and fertilizer irrigation action. When the root zone change value does not decrease, the heat and humidity change value increases, and the growth change value decreases, it is determined to be a green control action. When the root zone change value does not decrease, the heat and humidity change value decreases, and the growth change value increases, it is determined to be a shaping and pruning action, and an execution instruction is generated. The feedback update module is used to send the execution instructions corresponding to each location segment to the corresponding irrigation branch and environmental control zone for execution, and update the location segment sequence and execution instructions according to the IoT monitoring data after execution, and output the cultivation results for the next management cycle.

[0024] Working Principle: This solution organizes IoT monitoring data from various collection points within the park across two adjacent management cycles, categorizing them by collection location, collection time, and fruit tree location. It then calculates root zone changes, temperature and humidity changes, and growth changes for each collection point, further determining whether these changes are increasing, decreasing, or remaining constant. Subsequently, within the same irrigation and environmental control zones, it compares the direction of change at each collection point in adjacent order, merging locations with consistent changes into location segments. Based on the combination of changes at each collection point within a location segment, it determines whether water and fertilizer irrigation, green pest control, or pruning is more suitable, generating corresponding execution instructions. After execution, the monitoring data is reread, the location segments and execution instructions are updated, and the next management cycle begins. Thus, the entire process is not handled uniformly across the entire area, but rather managed separately based on the actual, continuously changing locations, allowing fruit tree cultivation and seed / seedling management to be continuously adjusted according to changes in the field conditions. For example, in a semi-enclosed greenhouse blueberry cultivation scenario in Shanghai, the areas near the vents and the central part of the greenhouse often have different conditions in terms of heat, humidity, and root zone water return. The system first reads the soil, environmental, and growth data for these locations separately, calculates the changes, and finds that a segment of the location shows decreased root zone changes, increased heat and humidity changes, and decreased growth changes. This segment is then grouped into the same location segment, and water and fertilizer irrigation instructions are issued. If another segment shows no decrease in root zone changes, increased heat and humidity changes, and decreased growth changes, a green control instruction is issued. After this round of execution is completed, the system looks at the new monitoring results and re-determines how to manage the next round. The same approach is used in open-field vineyards or in the scenarios of facility fruit trees such as cherries and blueberries and open-field economic forests. The system first looks at the actual changes in adjacent locations, and then decides what to do for each segment based on the changes, thus avoiding over-management in some areas and under-management in others.

[0025] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An Internet of Things-based intelligent fruit tree cultivation method, characterized in that, include: S1. Read the IoT monitoring data of each collection location in the park during the current management cycle and the previous management cycle, and align them according to the fruit tree locations corresponding to each collection location to generate location records; S2. Based on the location records of each collection point, calculate the root zone change value, heat and humidity change value, and growth change value of each collection point in the current management cycle relative to the previous management cycle, and generate change records. S3. Within the same irrigation zone and environmental control zone, compare the change records in the adjacent order of each collection location, and merge consecutive collection locations with the same root zone change direction, heat and humidity change direction and growth change direction into location segments to generate a location segment sequence. Read the change records of each collection location within the same irrigation zone and environmental control zone, form adjacent location pairs according to the adjacent order of each collection location, and compare the root zone change direction, heat and humidity change direction, and growth change direction of each adjacent location pair. Write a continuous mark when all three are consistent, write a check mark when only two of the three are consistent, and write a boundary mark in other cases to generate the initial result. For each adjacent position pair corresponding to the mark to be inspected, read the initial results of the previous adjacent position pair and the next adjacent position pair respectively. If both the previous and next adjacent position pairs are written with continuous marks, and the current inconsistency item is in the same direction in both the previous and next adjacent position pairs, rewrite the current mark to be inspected as a continuous mark; otherwise, rewrite it as a boundary mark and generate the verification result. Read the initial and verification results in the order of each collection location. If there are both continuous and boundary markers at the same adjacent location, retain the boundary marker; otherwise, retain the marker in the verification result. If only the initial result exists, retain the initial result and generate the final marking result. Read the final marking results in the order of each collection location. Merge the collection locations connected end to end by consecutive markings into candidate segments. Take the next collection location corresponding to the boundary marking as the starting position of the next candidate segment to generate a candidate segment sequence. For two candidate segments that are adjacent at the beginning and end in the candidate segment sequence, compare the change records of the last collection position of the previous candidate segment and the first collection position of the next candidate segment. If the change direction of the root region, the change direction of heat and humidity, and the change direction of growth are all consistent, merge them into the same location segment. Otherwise, keep them separate to generate a location segment sequence. S4. Based on the change records corresponding to each location segment, determine the cultivation actions for each location segment. When the root zone change value decreases, the heat and humidity change value increases, and the growth change value decreases, it is determined to be a water and fertilizer irrigation action. When the root zone change value does not decrease, the heat and humidity change value increases, and the growth change value decreases, it is determined to be a green control action. When the root zone change value does not decrease, the heat and humidity change value decreases, and the growth change value increases, it is determined to be a shaping and pruning action. Generate execution instructions. S5. Send the execution instructions corresponding to each location segment to the corresponding irrigation branch and environmental control zone for execution, and update the location segment sequence and execution instructions according to the IoT monitoring data after execution, and output the cultivation results for the next management cycle.

2. The intelligent fruit tree cultivation method based on the Internet of Things according to claim 1, characterized in that: S1 includes: S1-1. Read the IoT monitoring data of the current management cycle and the previous management cycle according to the collection location, and arrange them in ascending order according to the collection time to generate a collection sequence; S1-2. For the collection sequence at each collection location, match the IoT monitoring data of the current management cycle with the IoT monitoring data of the previous management cycle according to the corresponding category of IoT monitoring data, and write the unmatched IoT monitoring data into the missing marker to generate cycle alignment results. S1-3. Associate the period alignment results of each collection location with the corresponding fruit tree location to generate a location record.

3. The intelligent fruit tree cultivation method based on the Internet of Things according to claim 2, characterized in that: S2 includes: S2-1. Read the location records of each collection location, extract the root zone data, heat and humidity data and growth data corresponding to the same collection location in the current management cycle and the previous management cycle, and generate a periodic data group. S2-2. For the periodic data groups at each collection location, sort the current management cycle data and the previous management cycle data in ascending order of collection time, and pair them one by one in the order of arrangement to generate root zone pairing group, heat and humidity pairing group and growth pairing group. S2-3. For the root zone pairing group, heat and humidity pairing group and growth pairing group at each collection location, calculate the difference item by subtracting the data of the previous management cycle from the data of the current management cycle. Then, calculate the root zone change value, heat and humidity change value and growth change value by dividing the sum of the differences by the number of pairing items.

4. The intelligent fruit tree cultivation method based on the Internet of Things according to claim 3, characterized in that: S2 further includes: S2-4. Determine the root region change value that is greater than zero as the root region change direction increasing, the root region change value that is less than zero as the root region change direction decreasing, and the root region change value that is equal to zero as the root region change direction remaining unchanged. Values ​​of heat and humidity change greater than zero are defined as indicating an increasing direction of heat and humidity change; values ​​of heat and humidity change less than zero are defined as indicating a decreasing direction of heat and humidity change; and values ​​of heat and humidity change equal to zero are defined as indicating an unchanged direction of heat and humidity change. Growth change values ​​greater than zero are defined as increasing growth direction, growth change values ​​less than zero are defined as decreasing growth direction, and growth change values ​​equal to zero are defined as unchanged growth direction. The change record is composed of root zone change value, heat and humidity change value, growth change value, root zone change direction, heat and humidity change direction, and growth change direction.

5. The intelligent fruit tree cultivation method based on the Internet of Things according to claim 4, characterized in that: S4 includes: S4-1. Read the change records corresponding to each location segment, and extract the root zone change value, heat and humidity change value and growth change value of each collection location according to the location segment to generate segment change group; S4-2. For each segment of the fragment, determine the segment change group item by item. The sampling location where the root zone change value decreases, the heat and humidity change value increases, and the growth change value decreases is determined as the corresponding water and fertilizer irrigation action. The sampling location where the root zone change value does not decrease, the heat and humidity change value increases, and the growth change value decreases is determined as the corresponding green prevention and control action. The sampling location where the root zone change value does not decrease, the heat and humidity change value decreases, and the growth change value increases is determined as the corresponding shaping and pruning action. Generate the results corresponding to the sampling location.

6. The intelligent fruit tree cultivation method based on the Internet of Things according to claim 5, characterized in that: S4 further includes: S4-3. For each location segment, count the number of collection locations corresponding to water and fertilizer irrigation actions, green pest control actions, and shaping and pruning actions in the results. If the number of collection locations is greater than the number of collection locations of the other two, the action corresponding to the number of collection locations is determined as the cultivation action of that location segment. If the number of collection locations of two or three is equal, count the sum of the absolute values ​​of root zone change, heat and humidity change, and growth change for the corresponding collection locations. If the sum of one of these values ​​is greater than the sum of the other two, the action corresponding to this sum is determined as the cultivation action of that location segment. If the sums are equal, determine the cultivation action of that location segment in the order of water and fertilizer irrigation actions, green pest control actions, and shaping and pruning actions, and generate the action results. S4-4. Associate the action results of each location segment with the corresponding location segment, irrigation branch and environmental control zone to generate execution instructions.

7. The intelligent fruit tree cultivation method based on the Internet of Things according to claim 6, characterized in that: S5 includes: S5-1: Read the execution instructions corresponding to each location segment, and send them to the corresponding irrigation branch and environmental control zone for execution according to the location segment, and generate an execution record; S5-2. Read the IoT monitoring data corresponding to the execution record after execution, align it according to the fruit tree position corresponding to each collection location, regenerate the position record, change record and position segment sequence, and generate the updated result; S5-3. Based on the update results, redetermine the execution instructions corresponding to each position segment, and combine the updated position segment sequence and execution instructions to form the cultivation results for the next management cycle.

8. An Internet of Things-based intelligent fruit tree cultivation system for implementing the Internet of Things-based intelligent fruit tree cultivation method of any one of claims 1-7, characterized in that, include: The alignment generation module is used to read IoT monitoring data from each collection location in the park during the current management cycle and the previous management cycle, and align them according to the corresponding fruit tree locations to generate location records. The change calculation module calculates the root zone change value, heat and humidity change value, and growth change value of each collection location relative to the previous management cycle in the current management cycle, based on the location records of each collection location, and generates change records. The segmentation module is used to compare change records in the adjacent order of each collection location within the same irrigation zone and environmental control zone. It merges consecutive collection locations with the same root zone change direction, heat and humidity change direction, and growth change direction into location segments, generating a location segment sequence. The action determination module determines the cultivation action for each location segment based on the change records corresponding to each location segment. Specifically, when the root zone change value decreases, the heat and humidity change value increases, and the growth change value decreases, it is determined to be a water and fertilizer irrigation action. When the root zone change value does not decrease, the heat and humidity change value increases, and the growth change value decreases, it is determined to be a green control action. When the root zone change value does not decrease, the heat and humidity change value decreases, and the growth change value increases, it is determined to be a shaping and pruning action, and an execution instruction is generated. The feedback update module is used to send the execution instructions corresponding to each location segment to the corresponding irrigation branch and environmental control zone for execution, and update the location segment sequence and execution instructions according to the IoT monitoring data after execution, and output the cultivation results for the next management cycle.

Images