Mulberry tree planting mode adaptive matching system

Abstract

The application discloses a mulberry tree planting mode adaptive matching system, which comprises a park coordination node, a land end device, a field communication network and a planting rule library. Based on the mulberry tree planting mode library and the phenological window, the system retrieves a candidate mode according to microenvironment clustering, issues a parameter range, collects specific indexes such as leaf temperature and leaf area index, and updates the template according to the feedback variable closed loop. The field edge node identifies transplanting stress or soil water potential anomaly and triggers emergency rearrangement in linkage with ecological load; and in a weak network scene, supports combined reporting and idempotent processing. The system improves the survival rate and post-harvest recovery, and reduces the energy consumption and mode misfit risk.

Description

Technical Field

[0001] This invention relates to the field of agricultural and forestry informatization and precision cultivation, specifically an adaptive matching system for mulberry planting patterns. Background Technology

[0002] Mulberry trees, as the core raw material crop of the silkworm industry chain, have biological characteristics such as a significant seedling recovery period after transplanting, high frequency of leaf harvesting, short post-harvest recovery period, and sensitivity to changes in water and soil temperature. Existing digital systems in mulberry orchards mainly rely on general work orders and basic monitoring: the platform issues tasks based on experience templates, the terminals transmit results, and manual adjustments are made when anomalies occur. This approach has the following main pain points in the mulberry tree scenario:

[0003] Existing templates are mostly based on experience or single-regional experience, making it difficult to link them with the microenvironment of the plot (slope position, soil texture, wind and sunlight conditions, historical pests and diseases) and phenological windows (seedling establishment, greening, rapid growth period, post-harvest recovery), resulting in overly coarse configurations of row spacing, irrigation and drainage, and leaf harvesting intensity.

[0004] The monitoring items are mainly general environmental quantities, lacking mulberry-specific field indicators (such as leaf temperature / canopy temperature difference, leaf area index, bud sprouting rate, root vitality indicators, etc.) to quantitatively distinguish the quality of the model, making it difficult to support accurate template switching.

[0005] Anomalies are disconnected from the ecosystem: Although alarms can be triggered for abnormal equipment or water pressure, they are not linked to the ecological load (soil and water loss sensitivity, water quota limits, and surrounding ecological constraints), which can easily lead to suboptimal decisions such as excessive leaf harvesting / irrigation in order to rush the construction schedule. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the prior art and provide an adaptive matching system for mulberry planting patterns, including a park coordination node, a plot-end device, a field communication module, and a cloud data server; the cloud data server includes a planting rule base; the park coordination node, the plot-end device, and the cloud data server are respectively connected to the field communication module.

[0007] The planting rule base mentioned therein includes a mulberry planting pattern library, parameter templates, threshold and weight sub-libraries, phenological window configuration, ecological load coefficient configuration, and data dictionary;

[0008] The park coordination node is used to retrieve candidate planting patterns from the mulberry planting pattern library based on the plot microenvironment and the phenological window, and generate and issue an allocation order containing the pattern template identifier and parameter range to the plot end device.

[0009] The aforementioned field-end device is used to collect specific field indicators of mulberry trees within the phenological window and generate field records, and to report the field records and execution results to the park coordination node;

[0010] The park coordination node matches, scores, and updates the ranking of candidate planting patterns based on the reported feedback variables. After reaching the sampling coverage and verification cycle set by the rule base, it determines the target planting pattern and updates the parameter template and threshold on a rolling basis.

[0011] Preferably, the mulberry planting model library includes: row and plant spacing schemes, variety or rootstock-scion combinations, mulch / covering configuration, irrigation and drainage strategy templates, pruning and trunk setting rules, seedling establishment period management strategies, and upper limits for leaf harvesting intensity; the model entries are mapped to plot microenvironment zoning and phenological windows.

[0012] Preferably, the park coordination node performs micro-environment clustering on plots by using plot environmental data and soil type, topography, slope, sunlight distribution, wind conditions and historical pest and disease data to obtain plot cluster identifiers; the cluster identifiers participate in the pattern library retrieval and preliminary screening of candidate patterns.

[0013] Preferably, the process of retrieving candidate planting patterns from the mulberry planting pattern library based on the plot microenvironment and the phenological window includes: the phenological window includes the transplanting recovery period, the greening growth period, the rapid growth period of spring shoots, and the leaf harvesting recovery period; the pattern retrieval and parameter update are performed only within the task set that matches the current window, and parameter changes outside the window are postponed or enter a pending confirmation state.

[0014] Preferably, the mulberry-specific field indicators include leaf temperature or canopy temperature difference, leaf area index, bud sprouting rate, root vitality indicators, soil moisture / water potential and ground temperature; after the park coordination node performs quality verification and consistency checks on the mulberry-specific field indicators according to the data dictionary, it generates a set of matching elements for pattern evaluation.

[0015] Preferably, the feedback variables include the survival rate, the increase in plant height or diameter at breast height, the increase in leaf area index during the preset observation period, and the recovery time after leaf harvesting; when the feedback variables meet the improvement threshold and robustness conditions set by the rule base, the system improves the ranking of the corresponding mode item, otherwise it lowers or freezes its selectable state.

[0016] Preferably, it also includes abnormal ecological load linkage. When the field edge node identifies transplant stress or abnormal soil water potential and the ecological load coefficient of the target plot reaches the linkage condition set by the rule base, the park coordination node suppresses or restricts the leaf picking, topdressing or irrigation and drainage sub-tasks related to the plot and triggers a temporary switch of the pattern template mapped to the plot.

[0017] The beneficial effects of this invention are: the park coordination node, based on the micro-environment clustering of plots and the current phenological window retrieval of candidate modes, significantly reduces the probability of mismatch in the configuration of row spacing, cover, irrigation and drainage, and leaf harvesting intensity.

[0018] A matching element set was constructed using mulberry-specific field indicators such as leaf temperature / canopy temperature difference, leaf area index, bud sprouting rate, root vitality indicators, soil moisture / water potential and ground temperature, so that the evaluation of the model’s merits was transformed from experience-based judgment to data verification, which is particularly sensitive to the seedling establishment, greening and post-harvest recovery stages. Attached Figure Description

[0019] Figure 1 A schematic diagram illustrating the principle of an adaptive matching system for mulberry planting patterns;

[0020] Figure 2 A flowchart illustrating the process of cluster analysis of plots by coordinating nodes in the park, and using cluster identifiers for preliminary screening of planting patterns. Detailed Implementation

[0021] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings, but the scope of protection of the present invention is not limited to the following description.

[0022] The features and performance of the present invention will be further described in detail below with reference to embodiments.

[0023] like Figure 1 As shown, an adaptive matching system for mulberry planting patterns includes a park coordination node, a plot-end device, a field communication network, and a planting rule base.

[0024] The planting rule base includes a mulberry planting pattern library, parameter templates, threshold and weight sub-libraries, phenological window configuration, ecological load coefficient configuration, and data dictionary;

[0025] The park coordination node is configured to retrieve candidate planting patterns from the mulberry planting pattern library based on the plot microenvironment and the phenological window, and generate and issue an allocation order containing the pattern template identifier and parameter range to the plot end device.

[0026] The field-end device is configured to collect specific field indicators of mulberry trees within the phenological window and generate field records, and then report the field records and execution results to the park coordination node.

[0027] The park coordination node matches, scores, and updates the ranking of candidate planting patterns based on the reported feedback variables. After reaching the sampling coverage and verification cycle set by the rule base, the target planting pattern is determined and the parameter template and threshold are updated on a rolling basis.

[0028] The mulberry planting model library shall include at least the following fields: row and plant spacing scheme, variety or rootstock and scion combination, mulch / covering configuration, irrigation and drainage strategy template, pruning and trunk setting rules, seedling management strategy and leaf harvesting intensity limit; the model entries shall establish a mapping relationship between plot microenvironment zoning and phenological windows.

[0029] The park's coordination nodes use plot environmental data, soil type, topography, slope, sunlight distribution, wind conditions, and historical pest and disease data to perform micro-environment clustering on plots and obtain plot cluster identifiers; these cluster identifiers are used in the pattern library retrieval and preliminary screening of candidate patterns.

[0030] The phenological window includes at least the transplant recovery period, the greening and growth period, the rapid growth period of spring shoots, and the leaf harvesting recovery period. The system limits the execution of pattern retrieval and parameter updates to only within the task set that matches the current window. Parameter changes outside the window are postponed or enter a pending confirmation state.

[0031] The specific field indicators for mulberry trees include at least leaf temperature or canopy temperature difference, leaf area index, bud sprouting rate, root vitality indicators, soil moisture / water potential and ground temperature; after the park coordination node performs quality verification and consistency checks on the above indicators according to the data dictionary, it generates a set of matching elements for model evaluation.

[0032] The feedback variables include at least the survival rate, the increase in plant height or diameter at breast height, the increase in leaf area index during the preset observation period, and the recovery time after leaf harvesting. When the feedback variables meet the improvement threshold and robustness conditions set by the rule base, the system improves the ranking of the corresponding mode item; otherwise, it lowers or freezes its selectable state.

[0033] When the park coordination node generates an allocation order, it refines the parameter range into lightweight rules that can be issued to the plot side. This allows the plot-side devices to make fine-grained adjustments based on real-time observations without exceeding the upper limit constraints, and the fine-tuning records are sent back in the report for closed-loop learning.

[0034] Abnormal ecological load linkage is satisfied: When the field edge node identifies transplanting stress or abnormal soil water potential and the ecological load coefficient of the target plot (including soil and water loss sensitivity, water quota occupancy rate or adjacent ecological red line status) reaches the linkage conditions set by the rule base, the park coordination node will suppress or limit the leaf picking, topdressing or irrigation and drainage sub-tasks related to the plot and trigger a temporary switch of the pattern template mapped to the plot.

[0035] The field edge nodes and the park coordination nodes adopt an asymmetric division of labor: the field edge nodes only make local inferences about transplanting stress and soil water potential anomalies and form template switching suggestions, while the park coordination nodes perform global rearrangement and pattern determination across plots after aggregating data from multiple plots; when the conclusions of the two are inconsistent, the decision of the park coordination node shall prevail.

[0036] The field communication network supports low-bandwidth merged reporting: When the uplink quality is lower than the low-bandwidth threshold set by the rule base, the plot-end device merges the field record with the new round of task application or template switching suggestion in a single reporting transaction, and carries the idempotency flag and retransmission sequence number to ensure consistency.

[0037] The park's coordination nodes set priority coverage for emergency reordering, including validity and duration: within the validity period, emergency sorting overrides regular sorting; upon expiration or when the conditions for termination are met, the system reverts to regular sorting based on the most recent stability assessment results.

[0038] The system implements version control and audit backtracking for allocation orders, pattern entries, parameter templates, and data dictionaries, and performs deduplication and idempotency processing in re-application of tasks and anomaly checks to avoid statistical bias caused by duplicate requisition or evaluation.

[0039] Example 1: Adaptive matching of seedling establishment-greening period in hilly red soil mulberry orchards

[0040] Park Environment: Southwest hilly area with red soil, slope of 8–12°, strong annual winds, and significant diurnal temperature variations. The field communication network uses LoRa + intermittent 2G backhaul, resulting in large link fluctuations. One field-end device (including multi-source sensing and execution control) is deployed every 2–3 mu (approximately 0.16-0.2 hectares), and one field edge node is deployed every 30–50 mu (approximately 20–25 hectares). The edge nodes have local anomaly identification and buffered reporting capabilities.

[0041] The park's coordination nodes and planting rule base are centrally deployed in the park's computer room. The rule base includes a pattern library, thresholds and weights, phenological window configuration, ecological load coefficient configuration, and a data dictionary.

[0042] Model entry: M1 (seedling establishment and water retention type): row spacing 3.0 × 0.6 m; mulch film covering; drip irrigation with low flow rate and high frequency; leaf harvesting intensity limit prohibited; topdressing suspended.

[0043] M2 (Greening and Rooting Type): Row and plant spacing same as M1; medium flow and frequency of drip irrigation; trace amount of topdressing; leaves harvested before opening.

[0044] Phenological windows: W1 (transplanting and seedling establishment) 0–21 days; W2 (greening up) 22–45 days. Only maintenance tasks are allowed outside the windows; template switching is prohibited.

[0045] Field indicators: leaf temperature or canopy temperature difference, leaf area index, bud sprouting rate, root activity indicators, soil moisture / water potential, and soil temperature. Data is collected every 10 minutes at the field end; missing data is filled and boundary cleaning is performed upon reaching edge nodes; only data that passes quality verification is included in the matching feature set.

[0046] Feedback variables: survival rate, increase in plant height, increase in leaf area index during the 14-day observation period, and post-harvest recovery time (in this case, the plants were not opened during the seedling establishment / greening period, and only potential recovery capacity was recorded).

[0047] Microenvironment clustering: The platform divides mulberry orchards into three categories, A, B, and C, based on slope location, soil texture, wind exposure, and historical pest and disease data, and generates cluster labels.

[0048] Within W1, for Class A plots, candidate M1 and M2 are retrieved from the pattern library based on clustering identifiers. The platform generates an allocation order (including pattern template identifiers and parameter ranges) and distributes it to the plot-side devices. The plot-side devices fine-tune the drip irrigation frequency within the high-frequency range without exceeding the upper limit constraints and record the fine-tuning entries. Reports are submitted every 10 minutes when the link is normal; when the link is poor, low-bandwidth merging reporting is initiated. Before the end of W1, the platform statistically analyzes the compliance status of feedback variables. If Class A plots simultaneously meet the improvement threshold and robustness conditions in both survival rate and leaf area index improvement, the ranking of M1 in the Class A mapping is increased, and some parameters are fixed as the new template version. Upon entering W2, the platform re-retrieves and distributes the M2 parameter range, and the terminal releases the trace amounts of topdressing and the medium-frequency flow of drip irrigation, continuing the closed-loop evaluation.

[0049] Edge nodes only perform local inferences on soil water potential anomalies caused by transplanting stress and generate template switching suggestions. If an edge is identified as having abnormal soil water potential, and the ecological load coefficient of the plot (including slope erosion sensitivity and water quota occupancy) reaches the linkage condition, the platform triggers topdressing / irrigation and drainage flow restriction for the plot and maintains the M1 mode. The temporary coverage is valid for 48 hours and will be reviewed and restored after expiration.

[0050] When the uplink quality is lower than the low bandwidth threshold of the rule base, the land parcel terminal device merges a single reporting transaction: the latest field record + a second task application / template switching suggestion, and carries an idempotent flag and a retransmission sequence number; the platform side enables a deduplication strategy to ensure consistent statistical standards.

[0051] Example 2: Leaf Harvesting-Recovery Period Adaptation and Emergency Rearrangement in Sandy Soil Mulberry Orchards in River Valleys

[0052] Park environment: Inland river valley terrace, sandy loam soil, underground drip irrigation backbone network + small weather station; hot and dry summers with strong evapotranspiration. Communication is mainly 4G, with local obstructions causing frequent intermittent outages. One set of field-end devices is deployed for every 5 acres, with 2 field edge nodes deployed in each area; the edge nodes have local models, merged caching, and alarm reporting capabilities.

[0053] Model entry: M3 (conventional leaf-harvesting type): row spacing 2.8 × 0.7 m; moderate leaf-harvesting intensity; medium flow rate and medium frequency of drip irrigation; conventional topdressing.

[0054] M4 (Conservative Leaf Harvesting - High Temperature Period): Extended leaf harvesting interval; high frequency of drip irrigation with low to medium flow rate; reduced fertilizer application rate; thicker mulch.

[0055] Phenological windows: W3 (rapid spring shoot growth period) 46–90 days; W4 (leaf harvest recovery period) 91–150 days. This example covers cross-window operation from W3 to W4.

[0056] Field indicators: leaf area index, leaf temperature or canopy temperature difference, soil moisture / water potential, soil temperature, branch length and pest and disease indicators.

[0057] Feedback variables: plant height increase within W3, postharvest recovery time and leaf area index increase within W4.

[0058] Ecological load: The water quota occupancy rate and the ecological red line status of the river valley area are used as the sources of the ecological load coefficient.

[0059] The W3 platform uses microenvironment clustering results as an index to retrieve M3 / M4, generating allocation orders (including parameter ranges) and distributing them to the plot end. To adapt to local shading differences, the plot end device fine-tunes the leaf collection intensity of M3 within allowable limits and transmits the fine-tuning records back. The plot end collects field indicators every 10 minutes; the platform performs rolling calculations of sampling coverage and robustness every 30 minutes.

[0060] Three consecutive days of high temperatures occurred and abnormal soil water potential was detected at the edge nodes. At the same time, the ecological load coefficient reached the linkage condition. The edge generated template switching suggestion, and the platform temporarily switched the plot from M3 to M4 accordingly, and implemented emergency rearrangement of leaf harvesting, topdressing and irrigation and drainage tasks (valid for 72 hours).

[0061] When 4G is briefly congested, the terminal enables low-bandwidth merged reporting, which combines field records and handover suggestions into a single transaction report to avoid frame loss and duplicate statistics.

[0062] Before W4 ends, the platform analyzes the performance of feedback variables for the plot during and after the emergency period: if the post-harvest recovery time decreases and the leaf area index increases to the threshold and robustness condition, the ranking of M4 under the high temperature-low water potential scenario is upgraded; if the recovery after the period ends is average, M3 is restored and the optional state of M4 under this scenario is frozen for an observation period. The platform implements resource cascading constraints on adjacent plots, prioritizing water and fertilizer application and operation periods for plots undergoing emergency rearrangement, while low-priority tasks for other plots are postponed to avoid network-wide congestion.

[0063] Local inferences are made only for two patterns: transplant stress and abnormal soil water potential, to generate template switching suggestions; pest and disease indicators are only given hints and do not trigger template switching. After merging data from multiple plots, a global rearrangement and final pattern determination are performed; when the edge judgment is inconsistent with the platform's judgment, the platform's decision prevails, and the reasons for the conflict and backtracking evidence are recorded.

[0064] The above description is merely a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the concept described herein through the above teachings or related technologies or knowledge. Modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. An adaptive matching system for mulberry planting patterns, characterized in that, It includes a park coordination node, a plot-end device, a field communication module, and a cloud data server; the cloud data server includes a planting rule database; the park coordination node, the plot-end device, and the cloud data server are all connected to the field communication module. The planting rule base mentioned therein includes a mulberry planting pattern library, parameter templates, threshold and weight sub-libraries, phenological window configuration, ecological load coefficient configuration, and data dictionary; The park coordination node is used to retrieve candidate planting patterns from the mulberry planting pattern library based on the plot microenvironment and the phenological window, and generate and issue an allocation order containing the pattern template identifier and parameter range to the plot end device. The aforementioned field-end device is used to collect specific field indicators of mulberry trees within the phenological window and generate field records, and to report the field records and execution results to the park coordination node; The park coordination node matches, scores, and updates the ranking of candidate planting patterns based on the reported feedback variables. After reaching the sampling coverage and verification cycle set by the rule base, it determines the target planting pattern and updates the parameter template and threshold on a rolling basis.

2. The adaptive matching system for mulberry planting patterns according to claim 1, characterized in that, The mulberry planting model library includes: row and plant spacing schemes, variety or rootstock-scion combinations, mulch / covering configuration, irrigation and drainage strategy templates, pruning and trunk setting rules, seedling establishment period management strategies, and upper limits for leaf harvesting intensity; model entries are mapped to plot microenvironment zoning and phenological windows.

3. The adaptive matching system for mulberry planting patterns according to claim 1, characterized in that, The park's coordination nodes use plot environmental data, soil type, topography, slope, sunlight distribution, wind conditions, and historical pest and disease data to perform micro-environment clustering on plots and obtain plot cluster identifiers; these cluster identifiers are used in the pattern library retrieval and preliminary screening of candidate patterns.

4. The adaptive matching system for mulberry planting patterns according to claim 1, characterized in that, The process of retrieving candidate planting patterns from the mulberry planting pattern library based on the plot microenvironment and the phenological window includes: the phenological window includes the transplanting and seedling recovery period, the greening and growth period, the rapid growth period of spring shoots, and the leaf harvesting recovery period; the pattern retrieval and parameter update are only performed within the task set that matches the current window, and parameter changes outside the window are postponed or enter a pending confirmation state.

5. The adaptive matching system for mulberry planting patterns according to claim 1, characterized in that, The specific field indicators for mulberry trees include leaf temperature or canopy temperature difference, leaf area index, bud sprouting rate, root vitality indicators, soil moisture / water potential and ground temperature; after the park coordination node performs quality verification and consistency checks on the specific field indicators for mulberry trees according to the data dictionary, it generates a set of matching elements for model evaluation.

6. The adaptive matching system for mulberry planting patterns according to claim 1, characterized in that, The feedback variables include survival rate, increase in plant height or diameter at breast height, increase in leaf area index during the preset observation period, and recovery time after leaf harvesting. When a feedback variable meets the enhancement threshold and robustness conditions set by the rule base, the system promotes the ranking of the corresponding pattern entry; otherwise, it reduces or freezes its selectable state.

7. The adaptive matching system for mulberry planting patterns according to claim 1, characterized in that, It also includes abnormal ecological load linkage. When the field edge node identifies transplant stress or abnormal soil water potential and the ecological load coefficient of the target plot reaches the linkage conditions set by the rule base, the park coordination node will suppress or limit the leaf picking, topdressing or irrigation and drainage sub-tasks related to the plot and trigger a temporary switch of the pattern template mapped to the plot.

Images