A Precision Management Method for Tea Gardens Based on Multi-Source Data Fusion
By dividing the tea garden into multiple micro-zones, collecting multi-source data to construct a tea tree response chain, forming a growth response fingerprint, and dynamically generating management parameters, the problem of difficulty in reflecting differences in environmental change responses in tea garden management is solved, thereby improving the precision and adaptability of tea garden management.
Patent Information
- Application Number
- CN202611403000.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-10
- Publication Date
- 2026-11-10
AI Technical Summary
Existing tea garden management systems are unable to fully reflect the actual differences in the response of tea trees to environmental changes in different tea garden areas and at different growth stages. They lack continuous analysis of the dynamic response process of tea trees after environmental disturbances, and the judgment of the effect of management operations relies on a single environmental indicator, resulting in management lag or over-operation.
The tea garden is divided into multiple micro-zones, and multi-source environmental data and tea tree phenotypic data are collected. Environmental disturbance events are identified through a sliding time window, a tea tree response chain is constructed, a growth response fingerprint is formed, management parameters are dynamically generated, and a closed-loop management of environmental disturbance characteristics and tea tree responses is established.
It improves the precision and adaptability of tea garden management, reduces over-irrigation or under-fertilization caused by fixed threshold management, and achieves continuous optimization of tea garden management.
Abstract
Description
Technical Field
[0001] This application relates to the field of intelligent tea garden management technology, and in particular to a method for precise tea garden management based on multi-source data fusion. Background Technology
[0002] Existing tea garden management typically relies on single or limited environmental indicators such as soil moisture content, air temperature, and light intensity for irrigation, fertilization, and shading management. Although some management systems can collect multi-source environmental data, most still use fixed thresholds or preset rules for decision-making, making it difficult to fully reflect the actual differences in the response of tea trees to environmental changes in different tea garden areas and at different growth stages.
[0003] Meanwhile, existing tea garden management processes often lack continuous analysis of the dynamic response process of tea trees after environmental disturbances occur. The time correlation between environmental changes and tea tree phenotypic changes is difficult to utilize effectively. Furthermore, after management operations are completed, the effect is often judged only based on environmental indicators, and the actual response results of tea trees cannot be fed back to subsequent management parameters. Summary of the Invention
[0004] Therefore, it is necessary to provide a precise management method for tea gardens based on multi-source data fusion to address the aforementioned technical issues.
[0005] This application provides a tea garden precision management method based on multi-source data fusion, which adopts the following technical solution: Specific steps are as follows: S1. Divide the tea garden to be managed into multiple tea garden micro-regions, and obtain environmental data and tea tree phenotypic data corresponding to each tea garden micro-region; wherein, the environmental data includes at least two of soil moisture content, air temperature, air humidity, light intensity and soil nutrient status, and the tea tree phenotypic data includes at least two of canopy temperature, new shoot length, new shoot number, tea bud density and leaf phenotypic parameters; S2. Time-align the environmental data and tea tree phenotypic data corresponding to the same tea garden micro-region according to a unified time benchmark, and perform sliding time window processing on the continuously collected environmental data. Based on the change amplitude and duration of environmental parameters relative to the benchmark state within the sliding time window, identify environmental disturbance events and determine the occurrence time, disturbance type, disturbance amplitude and disturbance duration of the environmental disturbance events. S3. Taking the occurrence time of the environmental disturbance event as the response start time, extract the tea tree phenotypic data within a preset response time window after the occurrence of the environmental disturbance event, and sort the changes in the tea tree phenotypic data relative to the baseline state before the occurrence of the environmental disturbance by time to form a tea tree response chain corresponding to the environmental disturbance event; the tea tree response chain includes at least the response delay time, the magnitude of phenotypic change, and the response duration. S4. Classify the tea tree response chains formed in the same tea garden micro-region under different environmental disturbance events according to the disturbance type, and normalize the disturbance amplitude, disturbance duration, response delay time, phenotypic change amplitude and response duration to form growth response characteristics corresponding to the tea garden micro-region, and form the growth response fingerprint of the tea garden micro-region according to the combination relationship of multiple growth response characteristics in a continuous management cycle. S5. Obtain the growth response fingerprint corresponding to the current management cycle and compare it with the historical growth response fingerprint of the tea garden micro-region. Determine the state migration direction of the current tea garden micro-region relative to the historical growth state based on the direction and degree of change of response parameters under different environmental disturbance types. The state migration direction includes at least one of the following: stable state, stress state, and recovery state. S6. Based on the state transition direction, the type of environmental disturbance that caused the state transition, and the current environmental data, determine the management type corresponding to the tea garden micro-zone, and determine the management parameters based on the response results corresponding to the current environmental data and the historical effective management operations. S7. Perform at least one of the following management operations on the corresponding tea garden micro-region according to the management type and management parameters: water regulation, fertilizer regulation, light regulation or pest and disease intervention. After the management operation is completed, re-collect the environmental data and tea tree phenotypic data of the tea garden micro-region according to the sampling time benchmark corresponding to the management operation. S8. Based on the changes in response delay time, phenotypic change magnitude, and response duration in the corresponding tea tree response chain before and after the management operation, determine the response effect of this management operation, and associate and store the management type, management parameters, environmental data before the management operation, environmental data after the management operation, and response effect. S9. Based on the management parameters and their response effects corresponding to the effective management operations formed in multiple consecutive management cycles, update the growth response fingerprint of the tea garden micro-region, and establish the parameter correspondence between the current environmental disturbance characteristics, management parameters and tea tree response characteristics, so as to determine the management parameters for the next management cycle based on the updated growth response fingerprint and the parameter correspondence. Preferably, the tea garden micro-zones are initially divided based on at least one of tea tree variety, planting density, terrain height, slope aspect, soil type, and historical environmental data. After the initial division, the difference values of corresponding environmental parameters between adjacent tea garden areas are calculated. When the difference value exceeds the preset micro-zone boundary adjustment threshold, the corresponding area is re-divided into different tea garden micro-zones so that the environmental data within the same tea garden micro-zone has preset consistency.
[0006] Preferably, in step S2, the value of the environmental parameter at the start of the sliding time window is used as the reference value to calculate the maximum change range of the environmental parameter within the sliding time window, and whether an environmental disturbance has occurred is determined according to the relationship between the maximum change range and the corresponding preset change threshold; when the maximum change range reaches the corresponding preset change threshold and the duration reaches the preset duration threshold, the change corresponding to the environmental parameter is determined as an environmental disturbance event, and the disturbance type of the environmental disturbance event is determined according to the type of the environmental parameter.
[0007] Preferably, in step S3, the tea tree phenotypic baseline value within a preset baseline time window before the environmental disturbance event occurs and the tea tree phenotypic change value within a preset response time window after the environmental disturbance event occurs are obtained respectively. The time difference between the time when the tea tree phenotypic change is first detected to exceed the preset response threshold and the time when the environmental disturbance event occurs is determined as the response delay time. The difference between the tea tree phenotypic change value and the corresponding tea tree phenotypic baseline value is determined as the phenotypic change amplitude. The time when the phenotypic change value is continuously maintained above the preset response threshold is determined as the response duration.
[0008] Preferably, in step S4, the growth response fingerprint includes disturbance type features, disturbance amplitude features, disturbance duration features, response delay features, phenotypic change amplitude features, and response duration features corresponding to the tea garden micro-region; wherein, for the same disturbance type, the corresponding growth response features formed in multiple management cycles are combined in chronological order, and the dynamic growth response fingerprint corresponding to the tea garden micro-region is formed according to the change direction and degree of change between each growth response feature.
[0009] Preferably, in step S5, corresponding response direction determination rules are set for different types of environmental disturbances, and the changes in the corresponding response parameters in the growth response fingerprint of the current management cycle and the historical growth response fingerprint are calculated; when the change direction of at least two response parameters is consistent with the preset stress response direction under the corresponding environmental disturbance type, the tea garden micro-region is determined to be migrating to a stress state; when the change direction of at least two response parameters is opposite to the preset stress response direction, the tea garden micro-region is determined to be migrating to a recovery state; when the changes in the response parameters are all within the preset stable range, the tea garden micro-region is determined to be in a stable state.
[0010] Preferably, in step S6, when the tea garden micro-area migrates to a stress state, the management type is determined based on the type of environmental disturbance and the current environmental data; when the type of environmental disturbance is soil moisture change disturbance, it is determined to be a water regulation type; when the current soil nutrient status is lower than the corresponding nutrient threshold and the tea tree phenotypic data simultaneously shows response characteristics corresponding to nutrient deficiency, it is determined to be a fertilization regulation type; when the light intensity or canopy temperature reaches the corresponding light regulation threshold, it is determined to be a light regulation type; the management parameters are determined based on the current environmental data, the disturbance amplitude, and the response effects corresponding to historical effective management operations.
[0011] Preferably, in step S8, when at least two of the following conditions are met: the response delay time after the management operation is shorter than before the management operation, the phenotypic change magnitude changes in the direction corresponding to the stable state, and the response duration changes in the direction corresponding to the stable state, the corresponding management operation is marked as a valid management operation; when the validity conditions are not met, the corresponding management operation is marked as an inefficient management operation, and the management parameters for the next management cycle are adjusted according to the difference between the environmental disturbance magnitude corresponding to the inefficient management operation and the management parameters.
[0012] Preferably, when the same management type is used in the same tea garden micro-region for multiple consecutive management cycles and at least two effective management operations are formed, a correspondence between management parameters and tea tree response characteristics is established based on the environmental disturbance amplitude, management parameters and response effects corresponding to each effective management operation; in the next management cycle, the management parameters corresponding to the historical effective management operations are proportionally adjusted based on the difference between the current environmental disturbance amplitude and the historical environmental disturbance amplitude, and the adjusted management parameters are used as the initial management parameters for the current management cycle.
[0013] Preferably, after the tea garden micro-region completes the management operation, the updated growth response fingerprint is compared with the growth response fingerprint of the adjacent tea garden micro-region. When the two tea garden micro-regions have the same type of environmental disturbance, and at least two of the corresponding response delay time, phenotypic change amplitude, and response duration change in the same direction, and the difference between the corresponding response parameters is within the preset association threshold range, the two tea garden micro-regions are determined as associated response micro-regions. For the associated response micro-regions, the joint sampling frequency of environmental data and tea tree phenotypic data is increased in the next management cycle, and growth state migration judgment is performed separately.
[0014] In summary, this application includes the following beneficial technical effects: By constructing a correspondence between environmental disturbance events and tea tree response chains, and forming growth response fingerprints by considering response delay time, phenotypic change magnitude, and response duration, we can more accurately characterize the dynamic response differences of different tea garden micro-plots to environmental changes. By comparing current and historical growth response fingerprints to determine the direction of growth state migration, and combining them with historical effective management results to dynamically generate management parameters, we can reduce over-irrigation, under-fertilization, or management lag caused by fixed threshold management. At the same time, by updating the growth response fingerprints and management parameters through the response effects after management, we can form a closed loop of "monitoring-judgment-management-evaluation-updating" in tea garden management, thereby improving the accuracy, adaptability, and continuous optimization capabilities of tea garden micro-plot management. Detailed Implementation
[0015] The present invention will be further described below with reference to specific embodiments. It should be noted that the following embodiments are used to illustrate the technical solution of the present invention, and not to limit the scope of protection of the present invention. Those skilled in the art can make adaptive adjustments to the specific parameters, sampling period, and management method without departing from the technical concept of the present invention.
[0016] In this embodiment, a tea garden to be managed is taken as the management object. The tea garden is initially divided according to its topography, tea variety, planting density, and soil type. The initial division results are further corrected by combining historical environmental data, dividing the entire tea garden into multiple tea garden micro-regions with relatively independent environmental characteristics. Each tea garden micro-region is assigned a unique micro-region number, and a correspondence between the micro-region number and the geographical location is established, so that subsequent environmental data, tea tree phenotypic data, environmental disturbance events, and management operations can all be mapped to specific tea garden micro-regions.
[0017] Regarding environmental data collection, at least two environmental data points from soil moisture content, air temperature, air humidity, light intensity, and soil nutrient status are acquired for each tea garden micro-plot. Soil moisture content can be obtained using a soil moisture sensor, air temperature and humidity using temperature and humidity sensors, light intensity using a light sensor, and soil nutrient status using a soil nutrient detection device. In this embodiment, to improve the reliability of environmental disturbance identification, at least soil moisture content, air temperature, air humidity, and light intensity are collected simultaneously.
[0018] Tea tree phenotypic data is acquired through phenotypic acquisition devices installed in micro-plots of the tea garden, including at least two of the following: canopy temperature, shoot length, number of shoots, tea bud density, and leaf phenotypic parameters. In this embodiment, infrared temperature acquisition is used to acquire canopy temperature, and image acquisition devices are used to acquire images of the tea tree canopy, shoots, and tea buds. Phenotypic parameters such as shoot length, number of shoots, tea bud density, leaf color, and leaf area are extracted from the image data at continuous time intervals.
[0019] All environmental data and tea tree phenotypic data are tagged with a unified timestamp. For example, natural time is used as the unified time base, and each data collection records at least the micro-region number, collection time, data type, and corresponding data value, thereby ensuring that environmental data and tea tree phenotypic data can correspond according to the collection time.
[0020] For the division of tea garden micro-zones, after the initial division, environmental data from adjacent areas are further compared. Taking soil moisture content as an example, if the difference in soil moisture content between two adjacent areas is within a preset consistency range at multiple consecutive sampling times, and there are no significant differences in parameters such as canopy temperature and air temperature, the current micro-zone boundary can be maintained. However, if the differences in soil moisture content, canopy temperature, or other environmental parameters between adjacent areas exceed a preset micro-zone boundary adjustment threshold, then the two areas are considered to have different environmental response bases, and the corresponding areas are re-divided into different tea garden micro-zones. Through this method, relatively consistent environmental conditions are maintained within the same tea garden micro-zone, while areas with significant environmental differences are managed separately.
[0021] After completing the micro-region division, a sliding time window process is applied to the continuous environmental data of each tea garden micro-region. Specifically, a sliding time window of length T is established based on the continuously collected environmental data, where T can be set according to the actual collection cycle of the tea garden. Within each sliding time window, the average value of environmental parameters corresponding to the initial stage of the window or the previous stable time period is used as the baseline value, and the maximum change of environmental parameters relative to the baseline value within the current window is calculated.
[0022] Taking soil moisture content as an example, the variation range of soil moisture can be calculated in the following way: ,in, The current soil moisture content. For the soil moisture content corresponding to the baseline time period, This represents the range of change in soil moisture.
[0023] For air temperature, it can be calculated as follows: ,in, The current air temperature, The reference air temperature.
[0024] When the change in a certain environmental parameter reaches a preset threshold and this change persists for a preset duration, the change is identified as an environmental disturbance event. For example, when soil moisture content continuously decreases relative to the baseline and reaches a preset threshold, it is identified as a soil moisture change disturbance; when air temperature continuously rises and exceeds a corresponding temperature change threshold, it is identified as a temperature change disturbance; and when light intensity undergoes a sustained and significant change, it is identified as a light change disturbance.
[0025] For each environmental disturbance event, an independent disturbance event record is established, including the tea garden micro-region number, the time of occurrence of the disturbance, the type of disturbance, the magnitude of the disturbance, and the duration of the disturbance. For example, a tea garden micro-region at time... If a sustained decrease in soil moisture content occurs, record the following: ,in, Indicates the micro-region number of the tea garden. Indicates the time when the disturbance occurred. Indicates the type of disturbance. Indicates the magnitude of the disturbance. Indicates the duration of the disturbance.
[0026] After identifying the environmental disturbance event, the time of occurrence of the disturbance (t_0) is taken as the starting time of the tea tree response. A baseline time window before the disturbance and a response time window after the disturbance are established. The phenotypic data of the tea tree before the disturbance are used to determine the current baseline state of the tea tree, and the continuous phenotypic data after the disturbance are used to determine the actual response of the tea tree to the environmental disturbance.
[0027] For example, for canopy temperature, the average canopy temperature over a period of time prior to the occurrence of environmental disturbance is used as the baseline value for canopy temperature: ,in, The temperature is the canopy temperature collected for the i-th time before the disturbance occurs, and N is the number of data points within the baseline time window.
[0028] After the disturbance occurs, the changes in phenotypic parameters such as canopy temperature, shoot length, number of shoots, and tea bud density relative to the baseline state are calculated in chronological order. When the change in a certain phenotypic parameter first exceeds a preset response threshold, that moment is recorded as the response moment for that phenotypic parameter, and then... Calculate the response delay time for the corresponding phenotypic parameters, where, This is the moment when the phenotypic change is first detected to exceed the response threshold. The moment when the environmental disturbance occurs. This corresponds to the response delay time.
[0029] The magnitude of phenotypic variation can be determined based on: , to perform calculations, where These are the perturbed phenotypic parameters. This represents the baseline phenotypic value before the disturbance. For phenotypic parameters with different dimensions, further normalization can be used. ,in A very small positive number set to prevent the denominator from being zero.
[0030] When the phenotypic change consistently exceeds the corresponding response threshold, the time between the moment it first exceeds the threshold and the moment it recovers to within the threshold range is defined as the response duration. This forms a tea tree response chain for a single environmental disturbance event.
[0031] For example, for a high-temperature disturbance event, the response chain of a tea garden micro-region can be represented as: R E ={(τ C A C D C ),(τ L A L D L ),(τ B A B D B )}, where τ C A C D C τ represents the response delay, magnitude of change, and duration corresponding to canopy temperature, respectively; L A L D L These represent the response parameters corresponding to the new shoot length; τ B A B D B These represent the response parameters corresponding to the tea bud density.
[0032] Based on this, the tea tree response chains generated in the same tea garden microplot during different management cycles were classified according to the type of environmental disturbance. For example, response chains for soil moisture change disturbance, high temperature disturbance, and light intensity change disturbance were established separately. For each type of disturbance, the disturbance amplitude, disturbance duration, response delay time, phenotypic change amplitude, and response duration formed in multiple management cycles were normalized.
[0033] In this embodiment, range normalization can be used: ,in, For the current eigenvalue, and These are the maximum and minimum values from historical data, respectively. These are the normalized eigenvalues.
[0034] The normalized features are combined in a predetermined order to form the growth response feature vector: F=[Z E Z A Z D Z τ Z R Z RD ], where Z E Z represents the characteristic corresponding to the type of disturbance. A Z represents the amplitude characteristics of the disturbance.D Z represents the duration characteristic of the disturbance. τ Z represents the response delay characteristic. R Z represents the magnitude of phenotypic variation. RD This indicates the duration of the response.
[0035] For the same tea garden micro-plot, multiple growth response feature vectors are generated over multiple consecutive management cycles, and these vectors are combined in chronological order to form a growth response fingerprint with temporal evolution characteristics: G M ={F1,F2,...,F n}, where G M This represents the growth response fingerprint corresponding to the tea garden micro-region numbered M, F n This represents the growth response characteristics formed during the nth management cycle.
[0036] This growth response fingerprint does not simply record the state of tea trees at a certain moment, but rather records the dynamic response characteristics of tea trees when faced with environmental disturbances. Therefore, it can characterize the response differences of different tea garden micro-regions under the same environmental change conditions.
[0037] When assessing growth status, a growth response fingerprint generated during the current management cycle is obtained, and response parameters corresponding to historical growth response fingerprints are extracted and compared. Different response direction determination rules are established for different types of environmental disturbances. For example, for soil moisture reduction disturbances, increased canopy temperature, decreased tea bud density, and decreased shoot growth rate can be considered as response directions corresponding to increased stress; for high temperature disturbances, increased canopy temperature, increased phenotypic change amplitude, and prolonged response duration can be considered as response directions corresponding to increased stress.
[0038] The changes in the current response parameters compared to historical response parameters can be calculated as follows: It also determines whether the current change belongs to the direction of stress enhancement based on the stress response direction corresponding to the pre-established disturbance type.
[0039] When the change direction of at least two response parameters is consistent with the preset stress response direction under the corresponding environmental disturbance type, the tea garden micro-region is determined to be migrating to the stress state; when the change direction of at least two response parameters is opposite to the stress direction, the tea garden micro-region is determined to be migrating to the recovery state; when the change of each response parameter is within the preset stable range, the tea garden micro-region is determined to be in a stable state.
[0040] To further improve the continuity of state judgments, a state transition index can be set: ,in, W represents the number of response parameters involved in the state determination. i D represents the weight of the i-th response parameter.i This is the direction determination value for the corresponding response parameter. When D... i A positive value is taken when the direction is consistent with the preset stress direction, and a negative value is taken when the direction is opposite to the stress direction. The absolute value can also be adjusted according to the degree of response change. When the threshold of the stress state is exceeded, it is determined to be a migration to a stress state. When the value is below the recovery state threshold, it is determined to be a transition to the recovery state; when it is in between, it is determined to be a stable state.
[0041] After determining the direction of growth state migration in the tea garden micro-region, the management type is determined based on the type of environmental disturbance that caused the state migration and the current environmental data.
[0042] When the soil moisture content is consistently below the preset moisture threshold, and the corresponding tea tree response fingerprint shows a trend corresponding to water stress, the water regulation type is determined to be used; when the current soil nutrient status is below the corresponding nutrient threshold, and the tea tree shoot length, shoot quantity, or tea bud density shows response characteristics corresponding to nutrient deficiency, the fertilization regulation type is determined to be used; when the light intensity or canopy temperature exceeds the corresponding light regulation threshold, the light regulation type is determined to be used.
[0043] The management parameters are not fixed values, but are determined by considering the current environmental disturbance magnitude and the response effects of historically effective management operations. For example, for moisture regulation, the initial water replenishment can be determined based on the difference between the current soil moisture content and the target moisture content: Q0=K Q (W T -W C ), where Q0 is the initial water replenishment parameter, W T For the target soil moisture content, W C K represents the current soil moisture content. Q This is the water replenishment conversion coefficient.
[0044] Further refine the initial management parameters based on the actual response results of historical management operations: Q=Q0(1+λE H ), where Q is the corrected water replenishment parameter, E H This represents the evaluation value of management effectiveness under similar or identical historical environmental disturbance conditions, where λ is the management parameter correction coefficient.
[0045] Fertilization parameters and shading parameters can also be modified in a similar way, so that different tea garden micro-regions can adopt different management intensities according to their own historical response characteristics.
[0046] After performing management operations, environmental and tea tree phenotypic data are recollected at the same sampling timeframe as before the operations. For example, if tea tree phenotypic data were collected at fixed time intervals before the management operations, the corresponding data will continue to be collected at the same time intervals after the operations are completed, in order to reduce the deviation in response results caused by changes in the sampling period.
[0047] The re-collected data was used to reconstruct the tea tree response chain after the management operation, and the response delay time, phenotypic change magnitude, and response duration before and after the management operation were calculated respectively.
[0048] When at least two of the following conditions are met: the response delay time after a management operation is shortened, the magnitude of the phenotypic change changes in the direction corresponding to the steady state, and the response duration changes in the direction corresponding to the steady state, the management operation is marked as a valid management operation.
[0049] For example, if the duration of the abnormal canopy temperature response is shortened after water regulation, and the downward trend in tea bud density is suppressed, then the water regulation is considered to have produced an effective response. In this case, the environmental conditions, the magnitude of the environmental disturbance, the management type, the management parameters, and the tea tree response results corresponding to this management operation are stored together.
[0050] If the response parameters after a management operation do not meet the validity criteria, the management operation is marked as an inefficient management operation. For inefficient management operations, the difference between the corresponding environmental disturbance amplitude and the disturbance amplitude corresponding to historical effective management operations is analyzed, and the management parameters are adjusted in the next management cycle to avoid repeating ineffective management intensity in continuous management processes.
[0051] When the same management type is used in the same tea garden micro-plot for multiple consecutive management cycles and at least two effective management operations are performed, the environmental disturbance amplitude, management parameters, and tea tree response results corresponding to each effective management operation are correlated. For example, water regulation is performed under different soil moisture content decrease rates, and the corresponding water replenishment parameters and tea tree response results after management are recorded, thereby establishing a correspondence between environmental disturbance amplitude, water replenishment parameters, and tea tree response effect.
[0052] When a similar environmental disturbance occurs in the next management cycle, the current disturbance magnitude will be compared with the disturbance magnitude in historically effective management operations, and the historically effective management parameters will be adjusted proportionally based on the difference between the two. For example: ,in These are the initial management parameters for the next management cycle. For historical effective management parameters, This represents the current magnitude of environmental disturbance. This represents the magnitude of environmental disturbances corresponding to historically effective management operations. Adjust the coefficients for the parameters.
[0053] The above methods enable management parameters to change dynamically according to the degree of environmental disturbance and the actual response of the tea trees, rather than using fixed irrigation amounts, fertilization amounts, or shading levels.
[0054] After a management cycle is completed, the new tea tree response chain and its corresponding management effect are added to the historical growth response fingerprint, updating the original growth response fingerprint. The updated growth response fingerprint simultaneously records environmental disturbance characteristics, tea tree response characteristics, and response results after management operations, enabling it to reflect the latest response status of the tea garden micro-region during the actual management process.
[0055] In addition, the updated growth response fingerprints of adjacent tea garden microregions are compared. When two adjacent tea garden microregions experience the same type of environmental disturbance, and at least two of their response delay time, phenotypic change magnitude, and response duration have the same direction of change, and the difference between the corresponding response parameters is within a preset association threshold range, the two tea garden microregions are identified as associated response microregions.
[0056] For identified associated response micro-regions, their joint sampling frequency is increased in the next management cycle, and environmental disturbance identification, growth response fingerprint updates, and state transition judgments are performed separately. For example, if a tea garden micro-region first detects a significant high-temperature disturbance, and its adjacent associated response micro-regions have historically shown similar response characteristics to the same high-temperature disturbance, the monitoring frequency of both micro-regions is simultaneously increased in the next management cycle to predict their state transition trends before significant phenotypic changes in the tea trees actually occur.
[0057] Through the above methods, this embodiment forms a continuous management process from the collection of multi-source environmental data and tea tree phenotypic data, tea garden micro-region division, environmental disturbance identification, tea tree response chain construction, growth response fingerprint formation, growth state migration judgment, management parameter generation, management effect evaluation to growth response fingerprint update, thereby transforming tea garden management from static management based on a single environmental indicator to micro-region differentiated management based on the actual dynamic response of tea trees.
[0058] In another embodiment, environmental data may further include one or more of soil electrical conductivity, soil pH, wind speed, rainfall, and leaf surface humidity, while tea tree phenotypic data may include one or more of canopy coverage, leaf area index, leaf color characteristics, and lesion area ratio. By adding data from different sources, the correspondence between environmental disturbance events and tea tree response chains can be further enriched, but the core remains the extraction of tea tree response features triggered by environmental disturbances, and the use of growth response fingerprints formed over continuous management cycles for state transition judgment and management parameter updates.
[0059] In practical applications, management operations such as water regulation, fertilization regulation, and light regulation can be performed using drip irrigation, sprinkler irrigation, variable-rate fertilization, shading equipment, or other mature agricultural management equipment, depending on the actual tea garden equipment conditions. This invention does not impose specific limitations on these methods. Each management device is only used to execute management parameters generated based on the results of growth state migration. The core technology lies in using multi-source data to establish a correlation between environmental disturbances and tea tree responses, and continuously updating the growth response fingerprint and management parameters of the corresponding tea garden micro-region based on the response results after actual management.
[0060] Through the above implementation methods, different tea garden micro-regions can adopt different management strategies based on their own historical response characteristics when facing the same environmental disturbances, and the management parameters can be adjusted again based on the actual response after management, thereby improving the pertinence, continuity and adaptability of tea garden precision management.
Claims
1. A method for precise management of tea gardens based on multi-source data fusion, characterized in that: The specific steps are as follows: S1. Divide the tea garden to be managed into multiple tea garden micro-regions, and obtain environmental data and tea tree phenotypic data corresponding to each tea garden micro-region; wherein, the environmental data includes at least two of soil moisture content, air temperature, air humidity, light intensity and soil nutrient status, and the tea tree phenotypic data includes at least two of canopy temperature, new shoot length, new shoot number, tea bud density and leaf phenotypic parameters; S2. Time-align the environmental data and tea tree phenotypic data corresponding to the same tea garden micro-region according to a unified time benchmark, and perform sliding time window processing on the continuously collected environmental data. Based on the change amplitude and duration of environmental parameters relative to the benchmark state within the sliding time window, identify environmental disturbance events and determine the occurrence time, disturbance type, disturbance amplitude and disturbance duration of the environmental disturbance events. S3. Taking the occurrence time of the environmental disturbance event as the response start time, extract the tea tree phenotypic data within a preset response time window after the occurrence of the environmental disturbance event, and sort the changes in the tea tree phenotypic data relative to the baseline state before the occurrence of the environmental disturbance by time to form a tea tree response chain corresponding to the environmental disturbance event; the tea tree response chain includes at least the response delay time, the magnitude of phenotypic change, and the response duration. S4. Classify the tea tree response chains formed in the same tea garden micro-region under different environmental disturbance events according to the disturbance type, and normalize the disturbance amplitude, disturbance duration, response delay time, phenotypic change amplitude and response duration to form growth response characteristics corresponding to the tea garden micro-region, and form the growth response fingerprint of the tea garden micro-region according to the combination relationship of multiple growth response characteristics in a continuous management cycle. S5. Obtain the growth response fingerprint corresponding to the current management cycle and compare it with the historical growth response fingerprint of the tea garden micro-region. Determine the state migration direction of the current tea garden micro-region relative to the historical growth state based on the direction and degree of change of response parameters under different environmental disturbance types. The state migration direction includes at least one of the following: stable state, stress state, and recovery state. S6. Based on the state transition direction, the type of environmental disturbance that caused the state transition, and the current environmental data, determine the management type corresponding to the tea garden micro-zone, and determine the management parameters based on the response results corresponding to the current environmental data and the historical effective management operations. S7. Perform at least one of the following management operations on the corresponding tea garden micro-region according to the management type and management parameters: water regulation, fertilizer regulation, light regulation or pest and disease intervention. After the management operation is completed, re-collect the environmental data and tea tree phenotypic data of the tea garden micro-region according to the sampling time benchmark corresponding to the management operation. S8. Based on the changes in response delay time, phenotypic change magnitude, and response duration in the corresponding tea tree response chain before and after the management operation, determine the response effect of this management operation, and associate and store the management type, management parameters, environmental data before the management operation, environmental data after the management operation, and response effect. S9. Based on the management parameters and their response effects corresponding to the effective management operations formed in multiple consecutive management cycles, update the growth response fingerprint of the tea garden micro-region, and establish the parameter correspondence between the current environmental disturbance characteristics, management parameters and tea tree response characteristics, so as to determine the management parameters for the next management cycle based on the updated growth response fingerprint and the parameter correspondence.
2. The tea garden precision management method based on multi-source data fusion according to claim 1, characterized in that: The tea garden micro-zones are initially divided based on at least one of the following: tea tree variety, planting density, terrain height, slope aspect, soil type, and historical environmental data. After the initial division, the difference values of corresponding environmental parameters between adjacent tea garden areas are calculated. When the difference value exceeds the preset micro-zone boundary adjustment threshold, the corresponding area is re-divided into different tea garden micro-zones so that the environmental data within the same tea garden micro-zone has preset consistency.
3. The tea garden precision management method based on multi-source data fusion according to claim 1, characterized in that: In step S2, the value of the environmental parameter at the start of the sliding time window is used as the reference value to calculate the maximum change range of the environmental parameter within the sliding time window, and whether an environmental disturbance has occurred is determined according to the relationship between the maximum change range and the corresponding preset change threshold. When the maximum change range reaches the corresponding preset change threshold and the duration reaches the preset duration threshold, the change corresponding to the environmental parameter is determined as an environmental disturbance event, and the disturbance type of the environmental disturbance event is determined according to the type of the environmental parameter.
4. The tea garden precision management method based on multi-source data fusion according to claim 1, characterized in that: In step S3, the tea tree phenotypic baseline value within a preset baseline time window before the environmental disturbance event occurs and the tea tree phenotypic change value within a preset response time window after the environmental disturbance event occurs are obtained respectively. The time difference between the time when the tea tree phenotypic change is first detected to exceed the preset response threshold and the time when the environmental disturbance event occurs is determined as the response delay time. The difference between the tea tree phenotypic change value and the corresponding tea tree phenotypic baseline value is determined as the phenotypic change amplitude. The time when the phenotypic change value is continuously maintained above the preset response threshold is determined as the response duration.
5. The tea garden precision management method based on multi-source data fusion according to claim 1, characterized in that: In step S4, the growth response fingerprint includes disturbance type features, disturbance amplitude features, disturbance duration features, response delay features, phenotypic change amplitude features, and response duration features corresponding to the tea garden micro-region; wherein, for the same disturbance type, the corresponding growth response features formed in multiple management cycles are combined in chronological order, and the dynamic growth response fingerprint corresponding to the tea garden micro-region is formed according to the change direction and degree of change between each growth response feature.
6. The tea garden precision management method based on multi-source data fusion according to claim 1, characterized in that: In step S5, corresponding response direction determination rules are set for different types of environmental disturbances, and the change in the corresponding response parameters in the growth response fingerprint of the current management cycle and the historical growth response fingerprint is calculated; when the change direction of at least two response parameters is consistent with the preset stress response direction under the corresponding environmental disturbance type, the tea garden micro-area is determined to be migrating to a stress state. When the change direction of at least two response parameters is opposite to the preset stress response direction, the tea garden micro-region is determined to be migrating towards the recovery state; when the change in the response parameters is within the preset stable range, the tea garden micro-region is determined to be in a stable state.
7. The tea garden precision management method based on multi-source data fusion according to claim 1, characterized in that: In step S6, when the tea garden micro-area migrates to a stress state, the management type is determined based on the type of environmental disturbance and the current environmental data. When the type of environmental disturbance is soil moisture change disturbance, it is determined to be a water regulation type. When the current soil nutrient status is lower than the corresponding nutrient threshold and the tea tree phenotypic data simultaneously shows response characteristics corresponding to nutrient deficiency, it is determined to be a fertilization regulation type. When the light intensity or canopy temperature reaches the corresponding light regulation threshold, it is determined to be a light regulation type. The management parameters are determined based on the current environmental data, the disturbance amplitude, and the response effects corresponding to historical effective management operations.
8. The tea garden precision management method based on multi-source data fusion according to claim 1, characterized in that: In step S8, when at least two of the following conditions are met: the response delay time after the management operation is shortened compared to before the management operation, the magnitude of the phenotypic change changes in the direction corresponding to the stable state, and the response duration changes in the direction corresponding to the stable state, the corresponding management operation is marked as a valid management operation. When the validity conditions are not met, the corresponding management operation is marked as an inefficient management operation, and the management parameters for the next management cycle are adjusted according to the difference between the environmental disturbance magnitude corresponding to the inefficient management operation and the management parameters.
9. The tea garden precision management method based on multi-source data fusion according to claim 8, characterized in that: When the same management type is used in the same tea garden micro-region for multiple consecutive management cycles and at least two effective management operations are performed, a correspondence between management parameters and tea tree response characteristics is established based on the environmental disturbance amplitude, management parameters and response effects corresponding to each effective management operation. In the next management cycle, the management parameters corresponding to the historical effective management operations are proportionally adjusted based on the difference between the current environmental disturbance amplitude and the historical environmental disturbance amplitude, and the adjusted management parameters are used as the initial management parameters for the current management cycle.
10. The tea garden precision management method based on multi-source data fusion according to claim 1, characterized in that: After the tea garden micro-region completes the management operation, the updated growth response fingerprint is compared with the growth response fingerprint of the adjacent tea garden micro-region. When the two tea garden micro-regions have the same type of environmental disturbance, and at least two of the corresponding response delay time, phenotypic change magnitude and response duration change in the same direction, and the difference between the corresponding response parameters is within the preset association threshold range, the two tea garden micro-regions are determined as associated response micro-regions. For the aforementioned associated response micro-regions, the joint sampling frequency of environmental data and tea tree phenotypic data will be increased in the next management cycle, and growth state migration judgments will be performed separately.
