An intelligent pruning and water management collaborative control method and system for fruit tree cultivation

Abstract

The application provides a kind of intelligent pruning and water management synergic control method and system of fruit tree cultivation, belong to cultivation management technical field, this method includes: based on the thermal infrared image sequence of the leaf surface of different directions of fruit tree canopy, the transpiration rate change of each direction leaf is extracted, and the initial time of local water stress of corresponding direction leaf is identified;When the time difference of local water stress initial time between any two direction leaves reaches the preset stress threshold, a thinning pruning scheme is constructed;After completing thinning pruning according to the thinning pruning scheme, the global recovery slope in the preset period after pruning and the global attenuation slope in the stressed period before pruning are calculated from the transpiration rate change curve of the whole canopy of fruit tree to obtain the water compensation priority coefficient;Based on water compensation priority coefficient, irrigation control is executed.The technical scheme provided by the application can solve the problem of pruning and irrigation synergy failure caused by the deviation of canopy hydraulic zoning and whole plant average diagnosis.

Description

Technical Field

[0001] This application relates to the field of cultivation management technology, and more specifically, to a method and system for intelligent pruning and water management collaborative control in fruit tree cultivation. Background Technology

[0002] Current cultivation management is undergoing a transformation from traditional experience-based management to intelligent and precise control. In the pruning stage, traditional manual pruning relies on experience and is labor-intensive, while intelligent pruning, with the help of drone remote sensing, image recognition, and 3D reconstruction technology, can accurately perceive the canopy structure and plan pruning paths. In terms of water management, the Internet of Things-based intelligent water and fertilizer system monitors soil moisture in real time through sensors and combines it with crop transpiration evaporation models to achieve remote precision irrigation and on-demand fertilization, effectively addressing the problems of low efficiency and insufficient precision of traditional flood irrigation. Moreover, precision irrigation under intelligent control can save more water resources than traditional methods.

[0003] In existing collaborative control systems for intelligent pruning and water management, firstly, sensor networks and image recognition technology are used to collect real-time data on tree growth, soil moisture, and microclimate. Then, this data is analyzed against built-in decision models (such as crop transpiration and evaporation models and canopy structure models) to generate collaborative decision commands. These commands are then executed to ensure precise water and nutrient supply to the tree under optimal structural conditions. However, in fruit tree cultivation, due to the spatial heterogeneity of shoot growth potential and the non-uniform interception of light and heat resources in different areas of the canopy, leaf groups in different locations often exhibit varying local transpiration demands. This state is further complicated by the heterogeneous distribution of soil moisture in the root zone and the hydraulic conductivity of branches. The coupling of segmented differences can lead to asynchronous stress phenomena in canopy hydraulic zoning. Traditional methods rely on the average transpiration rate of the whole plant or stem sap flow obtained from a few sensors to determine the onset of water stress. When a local area has entered the initial stress stage with a sharp drop in local stomatal conductance due to high radiation, the overall average value may still show that the water content is acceptable. This masks the spatiotemporal asynchronous nature of water stress, causing the average-state-based diagnosis and the uniform pruning and irrigation control strategies to lose synergy in both time and space, resulting in irreversible photosynthetic damage and fruit quality decline. Therefore, how to solve the problem of synergistic failure of pruning and irrigation caused by the deviation between canopy hydraulic zoning and the average diagnosis of the whole plant has become a difficult problem for the industry. Summary of the Invention

[0004] This application provides a method and system for intelligent pruning and water management coordination control in fruit tree cultivation, which can solve the problem of pruning and irrigation coordination failure caused by the deviation between canopy hydraulic zoning and the average diagnosis of the whole tree.

[0005] In a first aspect, this application provides a method for the intelligent pruning and water management coordinated control of fruit tree cultivation, comprising the following steps:

[0006] Thermo-infrared image sequences of leaf surfaces at different locations in the canopy of fruit trees were obtained, and the transpiration rate variation curves of leaves at each location were extracted.

[0007] The initial moment of local water stress for leaves in each direction was identified based on the transpiration rate change curves of the leaves in each direction.

[0008] When the time difference between the initial moments of local water stress between any two directional leaves reaches the preset stress threshold, a thinning and pruning scheme is constructed based on the orientation of the initial moment of local water stress and the transpiration contribution of the fruit tree canopy branch group.

[0009] After the thinning and pruning is completed according to the aforementioned thinning and pruning scheme, the water compensation priority coefficient, which characterizes the soil's effective water compensation capacity for the canopy structure after pruning, is calculated by the global recovery slope of the fruit tree's full canopy transpiration rate change curve during a preset period after pruning and the global attenuation slope during the stress period before pruning.

[0010] When the water compensation priority coefficient is less than the preset priority threshold, irrigation control is performed according to the pre-set irrigation plan during the plant's water self-buffering period until the water compensation priority coefficient rises back to the preset priority threshold, at which point the normal irrigation decision is restored.

[0011] In some embodiments, acquiring thermal infrared image sequences of leaf surfaces at different locations within the fruit tree canopy and extracting transpiration rate variation curves from leaves at each location specifically includes:

[0012] A thermal infrared image sequence of the leaf surface at different positions of the fruit tree canopy was acquired using a thermal infrared imager, and the canopy orientation was marked and time stamped in the thermal infrared image sequence.

[0013] Based on the thermal infrared image sequence after aligning the canopy orientation markers with the timestamps, pixel-level continuous temperature sequences of leaves in each orientation are extracted by segmenting the leaf area and the background area by setting grayscale thresholds.

[0014] Based on the pixel-level continuous temperature sequence of leaves in each direction and the ambient air temperature, relative humidity and net radiation flux of the canopy collected by the micro-weather station, the temporal stomatal conductance of leaves in each direction is inverted based on the canopy energy balance equation.

[0015] The transpiration rate variation curves of the blades in each direction were constructed based on the temporal stomatal conductance of the blades in each direction.

[0016] In some embodiments, identifying the initial moment of local water stress for leaves in a given orientation based on the transpiration rate change curves of leaves in each orientation specifically includes:

[0017] The transpiration rate variation curves of the leaves in each direction are filtered and smoothed to obtain the noise-reduced transpiration rate variation curves of the leaves in each direction.

[0018] The first-order difference operation is performed on the denoised transpiration rate variation curves of the blades in each direction to obtain the first-order difference sequence of transpiration rate of the blades in each direction.

[0019] Extract the first inflection point from the first-order difference sequence of transpiration rate of leaves in each direction, where the value changes from positive to negative and remains negative for the next three consecutive sampling points. Mark the time corresponding to this inflection point as the initial moment of local water stress for the corresponding leaf.

[0020] In some embodiments, the transpiration contribution of the fruit tree canopy branch group is determined by the following steps:

[0021] The transpiration rate variation curves of leaves in each position of the fruit tree canopy before thinning and pruning are obtained. The transpiration rate variation curves of leaves in each position are integrated in the time domain to obtain the transpiration rate per unit area of ​​leaves in each position.

[0022] Based on the transpiration rate per unit area of ​​leaves in each direction and the leaf area index of the corresponding canopy branch group, the total transpiration water consumption of the canopy branch group in each direction is calculated.

[0023] The transpiration contribution of each canopy branch is obtained by dividing the total transpiration water consumption of each canopy branch by the sum of the total transpiration water consumption of all canopy branches.

[0024] In some embodiments, when the time difference between the initial moments of local water stress between any two directional leaves reaches a preset stress threshold, the pruning plan is constructed based on the orientation of the earliest initial moment of local water stress combined with the transpiration contribution of the fruit tree canopy branches. Specifically, this includes:

[0025] Compare the initial time of local water stress in any two directions of the leaf, calculate the time difference between the initial time of local water stress between each pair of directions, and when the time difference between the initial time of local water stress between any pair of leaves reaches the preset stress threshold, mark the direction where the initial time of local water stress occurs first as the priority stress direction.

[0026] The transpiration contribution of the canopy branch group in the priority stress direction is compared with the preset contribution threshold. If the transpiration contribution of the canopy branch group in the priority stress direction exceeds the preset contribution threshold, then the priority stress direction is determined as the target direction for thinning and pruning.

[0027] Based on the difference between the transpiration contribution of the canopy branch group in the target pruning direction and the average transpiration contribution of the canopy branch groups in all directions, the pruning amount ratio in the target pruning direction is determined according to a direct proportional relationship, and a pruning plan containing the pruning target pruning direction and the corresponding pruning amount ratio is generated.

[0028] In some embodiments, the global recovery slope of the fruit tree canopy transpiration rate change curve within a preset time period after pruning is determined by the following steps:

[0029] After the pruning was completed, the transpiration rate change curves of leaves in all directions of the fruit tree canopy were collected within a preset time period. The transpiration rate change curves of leaves in all directions were weighted and averaged according to the same time stamp to obtain the first transpiration rate change curve of the entire fruit tree canopy.

[0030] Linear regression fitting is performed on the first transpiration rate change curve, and the slope of the fitted straight line is extracted as the global recovery slope of the fruit tree full canopy transpiration rate change curve within a preset time period after pruning.

[0031] In some embodiments, the different orientations of the fruit tree canopy include due east, due west, due south, and due north, with the center of the fruit tree trunk as the orthogonal reference.

[0032] Secondly, this application provides an intelligent pruning and water management collaborative control system for fruit tree cultivation, used to execute an intelligent pruning and water management collaborative control method for fruit tree cultivation. The system includes:

[0033] The acquisition module is used to acquire thermal infrared image sequences of the leaf surface at different positions in the canopy of fruit trees and extract the transpiration rate change curves of the leaves at each position.

[0034] The processing module is used to identify the initial moment of local water stress for leaves in each direction based on the transpiration rate change curves of the leaves in each direction.

[0035] The processing module is also used to construct a thinning and pruning scheme based on the orientation of the first occurrence of local water stress and the transpiration contribution of the fruit tree canopy branch group when the time difference between the initial moments of local water stress between any two directional leaves reaches a preset stress threshold.

[0036] The processing module is also used to calculate, after the pruning is completed according to the pruning plan, a water compensation priority coefficient that characterizes the soil's effective water compensation capacity for the pruning canopy structure by using the global recovery slope of the fruit tree's full canopy transpiration rate change curve during a preset period after pruning and the global attenuation slope during the stress period before pruning.

[0037] The execution module is used to perform irrigation control according to the pre-set irrigation plan during the plant's water self-buffering period when the water compensation priority coefficient is less than the preset priority threshold, until the water compensation priority coefficient rises back to the preset priority threshold, and then resumes the normal irrigation decision.

[0038] Thirdly, this application provides a computer device, the computer device including a memory and a processor, the memory storing code, the processor being configured to acquire the code and execute the above-described intelligent pruning and water management collaborative control method for fruit tree cultivation.

[0039] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described intelligent pruning and water management collaborative control method for fruit tree cultivation.

[0040] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects:

[0041] The intelligent pruning and water management collaborative control method and system for fruit tree cultivation provided in this application first acquires thermal infrared image sequences of leaf surfaces at different locations within the fruit tree canopy and extracts transpiration rate change curves for leaves at each location. Second, based on the transpiration rate change curves of leaves at each location, the initial moment of local water stress for the corresponding leaf location is identified. Further, when the time difference between the initial moments of local water stress between any two locations reaches a preset stress threshold, a thinning pruning scheme is constructed based on the location of the earliest initial moment of local water stress combined with the transpiration contribution of the fruit tree canopy branches. Then, after thinning pruning is completed according to the pruning scheme, a water compensation priority coefficient, representing the soil's effective water compensation capacity for the post-pruning canopy structure, is calculated using the global recovery slope of the fruit tree's full canopy transpiration rate change curve within a preset time period after pruning and the global attenuation slope within the stress period before pruning. Finally, when the water compensation priority coefficient is less than a preset priority threshold, irrigation control is executed according to a pre-set irrigation scheme within the plant's water self-buffering period until the water compensation priority coefficient rises back to the preset priority threshold, at which point the normal irrigation decision is restored.

[0042] Therefore, this application can solve the problem of synergistic failure of pruning and irrigation caused by the deviation between canopy hydraulic zoning and whole-tree average diagnosis. First, by acquiring thermal infrared image sequences of leaf surfaces at different locations within the fruit tree canopy and extracting transpiration rate change curves for leaves at each location, the water consumption status of the fruit tree canopy can be upgraded from a single whole-tree average monitoring to dynamic perception at the location level with different spatial resolutions. This allows for the continuous quantification and capture of the spatial heterogeneity of transpiration rate within the canopy caused by factors such as uneven light distribution and local differences in root water absorption, providing a multi-channel synchronous data foundation for the early localization of water stress. Second, based on the transpiration rate change curves of leaves at each location... Identifying the initial moment of local water stress in leaves at corresponding locations allows the transformation of the transpiration rate trend inversion caused by a decrease in stomatal conductance into a time-stamped stress initiation point. This enables the detection of water imbalance in specific locations at the data level before observable wilting symptoms appear throughout the canopy, thus advancing stress detection from judging visible symptoms to identifying physiological temporal characteristics. Furthermore, when the time difference between the initial moments of local water stress between any two leaf locations reaches a preset stress threshold, a thinning pruning plan is constructed based on transpiration contribution. This approach uses the spatial asynchronicity of canopy water stress as a trigger condition for pruning decisions, only allowing pruning when the time difference between different locations entering stress is significant and prioritizes stress. Directional pruning is initiated only when the water consumption weight of the stressed area is sufficiently high. This avoids ineffective damage to photosynthetic organs in non-stressed areas caused by uniform pruning across the entire canopy due to overall soil drought. This ensures that the selection of pruning location and intensity simultaneously considers stress relief efficiency and the maintenance of overall canopy yield. Then, after thinning pruning, a water compensation priority coefficient is calculated using the global recovery slope and global decay slope. This establishes a quantitative correlation between pruning as a physical control mechanism and the physiological response of soil moisture through the bidirectional dynamic changes in canopy transpiration rate. The relative relationship between recovery and decay rates replaces the single absolute value of soil moisture content to determine irrigation demand, thereby accurately distinguishing between pruning and non-stressed areas. The study examines whether the reduced canopy load after pruning has been adequately compensated by the soil's own effective moisture. Finally, when the water compensation priority coefficient is less than the preset priority threshold, a plant water self-buffering irrigation scheme is implemented until it recovers to the threshold, after which regular irrigation is resumed. This constructs an irrigation closed-loop control system with the canopy transpiration recovery rate as the feedback signal. High-frequency, low-quota compensatory irrigation is triggered only when the soil moisture compensation capacity is insufficient. Once the water compensation reaches the target, the system returns to normal irrigation, avoiding deep seepage and ineffective water consumption caused by blindly increasing irrigation after pruning. In summary, the technical solution provided in this application can solve the problem of synergistic failure of pruning and irrigation caused by the deviation between canopy hydraulic zoning and the overall plant average diagnosis. Attached Figure Description

[0043] Figure 1 This is an exemplary flowchart of a method for intelligent pruning and water management collaborative control of fruit tree cultivation according to some embodiments of this application;

[0044] Figure 2 This is a schematic diagram showing the positional distribution of thermal infrared images of the surface of fruit tree canopy leaves acquired from different orientations by an infrared imager according to some embodiments of this application.

[0045] Figure 3 This is an exemplary flowchart illustrating the determination of the initial moment of local moisture stress according to some embodiments of this application;

[0046] Figure 4 This is a schematic diagram of the structure of an intelligent pruning and water management collaborative control system for fruit tree cultivation, as shown in some embodiments of this application.

[0047] Figure 5 This is a schematic diagram of the structure of a computer device for implementing a method for intelligent pruning and water management collaborative control of fruit tree cultivation, according to some embodiments of this application. Detailed Implementation

[0048] To better understand the technical solution of this application, the technical solution of this application will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0049] refer to Figure 1 The figure is an exemplary flowchart of a method for intelligent pruning and water management collaborative control in fruit tree cultivation according to some embodiments of this application. The figure mainly includes the following steps:

[0050] In step S101, thermal infrared image sequences of leaf surfaces at different locations in the fruit tree canopy are obtained, and transpiration rate variation curves of leaves at each location are extracted.

[0051] It should be noted that, in this application, the fruit tree canopy refers to the three-dimensional spatial aggregate of the above-ground part of the fruit tree, which consists of the trunk, branches at all levels, and leaves, flowers, fruits, etc. growing on it. It is a dynamic system with biomass and voids with specific geometric shape, leaf area density, and microclimate characteristics. The environmental factors such as light, temperature, humidity, and wind inside it show significant vertical gradients and directional differences, and it is the main place for fruit trees to carry out photosynthesis, transpiration, and yield formation.

[0052] In some embodiments, acquiring thermal infrared image sequences of leaf surfaces at different locations within the fruit tree canopy and extracting transpiration rate variation curves for leaves at each location is achieved through the following steps:

[0053] A thermal infrared image sequence of the leaf surface at different positions of the fruit tree canopy was acquired using a thermal infrared imager, and the canopy orientation was marked and time stamped in the thermal infrared image sequence.

[0054] Based on the thermal infrared image sequence after aligning the canopy orientation markers with the timestamps, pixel-level continuous temperature sequences of leaves in each orientation are extracted by segmenting the leaf area and the background area by setting grayscale thresholds.

[0055] Based on the pixel-level continuous temperature sequence of leaves in each direction and the ambient air temperature, relative humidity and net radiation flux of the canopy collected by the micro-weather station, the temporal stomatal conductance of leaves in each direction is inverted based on the canopy energy balance equation.

[0056] The transpiration rate variation curves of the blades in each direction were constructed based on the temporal stomatal conductance of the blades in each direction.

[0057] In specific implementation, firstly, thermal infrared images of the leaf surfaces of the fruit tree canopy are acquired from the east, south, west, and north directions around the canopy using a thermal infrared imager, resulting in a sequence of thermal infrared images of the leaf surfaces at different locations within the canopy. These different locations include due east, due west, due south, and due north, with the center of the fruit tree trunk as the orthogonal reference. The corresponding azimuth code and timestamp are embedded into the metadata of each frame of the thermal infrared image sequence. The image timestamp is then aligned with the acquisition location to form a thermal infrared image sequence with canopy azimuth markers and timestamps aligned. Secondly, for the thermal infrared image sequence with canopy azimuth markers and timestamps aligned, [further steps are taken]. The Otsu's method is used to automatically calculate the grayscale segmentation threshold between the leaf region and the sky or soil background. Regions with pixel grayscale values ​​higher than this threshold are identified as leaf regions. The temperature values ​​of all pixels within each leaf region are extracted frame by frame and arranged chronologically to form a pixel-level continuous temperature sequence for leaves in each direction. This pixel-level continuous temperature sequence refers to the temperature change dataset of target pixels on the leaf surface in the same direction along the time axis. Then, micro-weather stations deployed near the fruit tree canopy collect data on the ambient air temperature, relative humidity, and net radiation flux of the canopy environment. The net radiation flux refers to the net balance of shortwave and longwave solar radiation received per unit area at the canopy height. Then, the pixel-level continuous temperature sequence of leaves in each direction, along with the canopy ambient air temperature, relative humidity, and net radiation flux collected by the micro-weather station, are input into the canopy energy balance equation for time-by-time inversion. The canopy energy balance equation is in the form that net radiation flux equals the sum of sensible heat flux, latent heat flux, and soil heat flux. Among these, soil heat flux is usually small and can be ignored. Sensible heat flux is obtained from the difference between leaf temperature and air temperature and aerodynamic impedance. Latent heat flux is related to the water vapor pressure difference and the overall canopy impedance. By introducing stomatal conductance as a variable to be solved into the equation and using the empirical relationship between aerodynamic impedance and leaf characteristic length for parameterization, the equation is solved time-by-time. The temporal stomatal conductance of leaves in each direction is calculated. The temporal stomatal conductance refers to the time series of the conduction capacity of leaf stomata for water vapor diffusion at each moment. Finally, the temporal stomatal conductance of leaves in each direction, along with the air temperature, relative humidity, and net radiation flux collected synchronously by the micro-weather station, are substituted into the Penman-Monteith equation. The equation is fully expressed as the transpiration rate equal to the sum of the net radiation flux term and the aerodynamic term divided by the combination function of the latent heat equivalent and the stomatal conductance term. After time-by-time calculation, the transpiration rate variation curves of leaves in each direction are output. The Penman-Monteith equation is a method used to calculate the transpiration and evaporation rate of vegetation, which will not be elaborated here.

[0058] refer to Figure 2This figure is a schematic diagram of the location distribution of thermal infrared images of the leaf surface of a fruit tree canopy collected from different directions by an infrared imager according to some embodiments of this application. A fruit tree with a trunk and complete leaves is drawn in the center of the figure. The circular dashed outline represents the outer boundary of the fruit tree canopy. With the center of the fruit tree trunk as the orthogonal reference, the east, west, south and north collection points equipped with thermal infrared imagers are set in the due east, due west, due south and due north directions, respectively. The lens of the infrared imager is horizontally aimed at the leaf surface of the corresponding direction of the canopy to collect thermal infrared images, completely covering the branch and leaf areas of different directions outside the canopy. It intuitively shows the uniform distribution position and shooting orientation of the thermal infrared imager in different directions outside the fruit tree canopy.

[0059] It should be noted that the transpiration rate change curve refers to the dynamic change trajectory of water transpiration rate per unit area of ​​leaf over time. In fruit tree cultivation, transpiration rate is a core physiological indicator connecting soil water supply, plant water conduction, and atmospheric evaporation demand in the canopy. Its time-series changes directly reflect the dynamic response of leaf stomata to water deficit. When local root water absorption is restricted or light and heat distribution within the canopy is uneven, the transpiration rate change curve in the corresponding location will show a trend reversal before the overall tree. By continuously monitoring the transpiration rate change curves in each location, the implicit water stress process can be transformed into a quantitative signal with temporal and spatial resolution, thereby providing continuous and comparable physiological data support for the construction of pruning schemes and the dynamic adjustment of subsequent irrigation strategies.

[0060] In step S102, the initial moment of local water stress for the corresponding leaf is identified based on the transpiration rate change curves of the leaves in each direction.

[0061] In some embodiments, reference Figure 3 As shown in the figure, this is an exemplary flowchart illustrating the determination of the initial moment of local water stress according to some embodiments of this application. In this embodiment, the identification of the initial moment of local water stress of leaves in corresponding directions based on the transpiration rate change curves of leaves in each direction is achieved by the following steps:

[0062] In step S1021, the transpiration rate variation curves of the blades in each direction are filtered and smoothed to obtain the noise-reduced transpiration rate variation curves of the blades in each direction.

[0063] In step S1022, a first-order difference operation is performed on the noise-reduced transpiration rate variation curves of the blades in each direction to obtain a first-order difference sequence of the transpiration rate of the blades in each direction.

[0064] In step S1023, the first turning point from positive to negative and the subsequent three consecutive sampling points remain negative are extracted from the first-order difference sequence of transpiration rate of leaves in each direction. The time corresponding to this turning point is marked as the initial time of local water stress of the corresponding leaf.

[0065] Specifically, the transpiration rate variation curves of the blades in each direction are input into the Savitzky-Golay filter. This filter achieves smoothing by performing polynomial least squares fitting of a specified order on local data points within a moving window. The window width is set to 5 sampling points and the polynomial order is 2. The window is slid point by point and the fitting value of the center point is calculated. This filters out high-frequency fluctuation components caused by instantaneous airflow disturbances or sensor noise, resulting in denoised transpiration rate variation curves of the blades in each direction that retain the overall trend of transpiration rate variation. The denoised transpiration rate variation curve refers to the smooth data sequence that reflects the temporal variation law of the blade transpiration rate after filtering out random noise.

[0066] Specifically, starting from the beginning of the first-order difference sequence of transpiration rates of leaves in each direction, the system iterates backward to search for the first inflection point where the value changes from positive to negative. The first-order difference values ​​of the three consecutive sampling points after this inflection point are then used to determine the sign of the values. If the first-order difference values ​​of these three consecutive sampling points remain negative, the inflection point is confirmed as the starting point of the monotonically decreasing transpiration rate. The sampling time corresponding to this inflection point is marked as the initial time of local water stress for the corresponding leaf. If any of the three consecutive sampling points after the inflection point have a positive value, the inflection point is discarded and the search continues backward until the conditions of a positive value turning into a negative value and the subsequent three consecutive sampling points remaining negative are met.

[0067] It should be noted that, in this application, the initial moment of local water stress refers to the starting point when the transpiration rate of leaves in different orientations changes from an upward trend to a continuous downward trend. This is used to characterize the starting point when leaves in a specific orientation of the fruit tree canopy transition from a normal transpiration state to a water stress state. When the root system's water supply capacity to a certain orientation begins to fall below the atmospheric evaporation demand of the leaves in that orientation, the stomata of the leaves will gradually change from an open state to a partially closed state to reduce the rate of water loss through transpiration. This stomatal regulation behavior is manifested on the transpiration rate change curve as a clear turning point from an upward trend to a continuous downward trend. The timing of this turning point can be located from a data perspective by determining the positive to negative value of the first-order difference sequence. Therefore, determining the initial moment of local water stress is to transform the early, implicit local water stress process, which has not yet led to overall canopy wilting, into a quantitative time signal with orientation resolution. This provides a direct basis for subsequent comparison of the time difference of the initial moments of local water stress between different orientations and for determining the priority stress orientation, thereby supporting the construction of a pruning scheme.

[0068] In step S103, when the time difference between the initial moments of local water stress between any two directional leaves reaches a preset stress threshold, a thinning and pruning scheme is constructed based on the orientation of the initial moment of local water stress and the transpiration contribution of the fruit tree canopy branch group.

[0069] In some embodiments, the transpiration contribution of the fruit tree canopy branch group is determined by the following steps:

[0070] The transpiration rate variation curves of leaves in each position of the fruit tree canopy before thinning and pruning are obtained. The transpiration rate variation curves of leaves in each position are integrated in the time domain to obtain the transpiration rate per unit area of ​​leaves in each position.

[0071] Based on the transpiration rate per unit area of ​​leaves in each direction and the leaf area index of the corresponding canopy branch group, the total transpiration water consumption of the canopy branch group in each direction is calculated.

[0072] The transpiration contribution of each canopy branch is obtained by dividing the total transpiration water consumption of each canopy branch by the sum of the total transpiration water consumption of all canopy branches.

[0073] It should be noted that, in this embodiment, the transpiration per unit area refers to the total amount of water lost by a unit leaf area in different orientations through transpiration during the irrigation cycle.

[0074] Specifically, based on the transpiration rate per unit area of ​​leaves in each direction and the leaf area index of the corresponding canopy branch group, the total transpiration water consumption of the canopy branch group in each direction is calculated. That is, the leaf area index of the canopy branch group in each direction is measured using a canopy analyzer. The leaf area index is the ratio of the sum of the single-sided area of ​​all leaves in the canopy branch group in each direction per unit ground area to the projected area of ​​the ground. The transpiration rate per unit area of ​​leaves in each direction is multiplied by the leaf area index of the corresponding canopy branch group to obtain the total transpiration water consumption of the canopy branch group in each direction. The total transpiration water consumption refers to the total amount of water actually lost by the canopy branch group in that direction through leaf transpiration during the entire irrigation cycle.

[0075] Specifically, the transpiration water consumption of each canopy branch group is divided by the sum of the transpiration water consumption of all canopy branch groups in all directions to obtain the transpiration contribution of each canopy branch group. That is, the total transpiration water consumption of each canopy branch group in all directions is normalized by dividing the total transpiration water consumption of each canopy branch group in each direction by the sum of the total transpiration water consumption of all canopy branch groups in all directions to obtain the transpiration contribution of each canopy branch group in each direction. The transpiration contribution refers to the proportion of water consumption of canopy branch groups in different directions to the total water consumption of the canopy. Its value ranges from 0 to 1 and the sum of the transpiration contribution of all directions is always equal to 1.

[0076] It should be noted that the reason for determining the transpiration contribution is that the water consumption capacity of branch groups in different directions of the fruit tree canopy is spatially heterogeneous due to differences in light conditions, leaf area density, and branch distribution. If pruning is carried out based solely on the order of stress occurrence and ignoring the transpiration weight of each branch group, it may lead to excessive removal of branch groups that contribute significantly to overall yield but respond to stress earlier, or insufficient removal to alleviate water imbalance. Therefore, determining the transpiration contribution quantifies the water consumption capacity of branch groups in each direction of the canopy into a comparable proportional value. When a certain direction is identified as the priority stress direction, the comparison between the transpiration contribution and the preset contribution threshold can determine whether the water consumption of the branch group in that direction has reached the level that requires pruning intervention. At the same time, the difference between the contribution and the canopy mean can be directly proportionally used to determine the proportion of thinning, thereby ensuring that the thinning pruning plan minimizes damage to the overall photosynthetic production capacity of the canopy while alleviating local water stress.

[0077] In some embodiments, when the time difference between the initial moments of local water stress between any two directional leaves reaches a preset stress threshold, the following steps are used to construct a thinning pruning scheme based on the orientation of the earliest initial moment of local water stress combined with the transpiration contribution of the fruit tree canopy branch groups:

[0078] Compare the initial time of local water stress in any two directions of the leaf, calculate the time difference between the initial time of local water stress between each pair of directions, and when the time difference between the initial time of local water stress between any pair of leaves reaches the preset stress threshold, mark the direction where the initial time of local water stress occurs first as the priority stress direction.

[0079] The transpiration contribution of the canopy branch group in the priority stress direction is compared with the preset contribution threshold. If the transpiration contribution of the canopy branch group in the priority stress direction exceeds the preset contribution threshold, then the priority stress direction is determined as the target direction for thinning and pruning.

[0080] Based on the difference between the transpiration contribution of the canopy branch group in the target pruning direction and the average transpiration contribution of the canopy branch groups in all directions, the pruning amount ratio in the target pruning direction is determined according to a direct proportional relationship, and a pruning plan containing the pruning target pruning direction and the corresponding pruning amount ratio is generated.

[0081] In practice, firstly, the initial times of local water stress for leaves in different orientations are paired according to their orientation labels. For each pair of orientations, the absolute difference in the initial times of local water stress is calculated as the time difference between the initial times of local water stress between the two orientations. The time difference refers to the deviation interval between the starting times of two leaves in different orientations entering a state of water stress. Each calculated time difference is then compared numerically with a preset stress threshold. The preset stress threshold is a time critical value set based on the characteristics of the fruit tree variety and cultivation experience, for example, 30 minutes. When the time difference between the initial times of local water stress between any pair of orientations first reaches or exceeds 30 minutes, the stress threshold is set from that orientation. The location with the lowest initial water stress among the leaves was selected and marked as the priority stress location. The priority stress location refers to the location in the entire canopy that first exhibits a continuous decrease in transpiration rate and enters a water stress state. Next, the transpiration contribution of the canopy branches at the priority stress location was compared with a preset contribution threshold. The preset contribution threshold is a proportional critical value set based on the reference lower limit of the contribution of a single location under uniform water consumption conditions in all directions of the canopy. For example, if the lower limit of the transpiration contribution of a single location is set to 15%, if the transpiration contribution of the canopy branches at the priority stress location is greater than 15%, it indicates that the water consumption of the branches at that location has reached a level that significantly affects the overall water balance of the canopy. If the level of influence is determined, the priority stress direction is formally designated as the target direction for thinning pruning. If the transpiration contribution of the canopy branch group in the priority stress direction does not exceed the preset contribution threshold, it indicates that the water consumption level of the branch group in that direction has not reached the level of significantly affecting the overall water balance of the canopy. In this case, the priority stress direction is not designated as the target direction for thinning pruning. The time difference of the initial local water stress time of the remaining direction combinations continues to be monitored and compared until a direction that meets the preset stress threshold condition and the transpiration contribution exceeds the preset contribution threshold is found. This is to avoid ineffective pruning of low contribution branch groups and unnecessary damage to the canopy. The target direction for thinning pruning refers to the direction in which some branches need to be removed in subsequent pruning operations. The canopy orientation is chosen to reduce water competition load. Finally, the arithmetic mean of the transpiration contribution of canopy branch groups in all orientations is calculated as the mean contribution value. The mean contribution value is obtained by subtracting this mean contribution value from the transpiration contribution of the canopy branch groups in the target orientation of the thinning pruning. The contribution deviation value is then substituted into a preset direct proportional mapping relationship to determine the thinning ratio. The direct proportional mapping relationship is in the form that the thinning ratio is equal to the contribution deviation value multiplied by a proportional coefficient. The proportional coefficient is pre-calibrated based on tree species pruning intensity tests. For example, if the proportional coefficient is set to 2, and the contribution deviation value is 0.1, then the thinning ratio is 20%. This generates a thinning pruning scheme that includes the target orientation of the thinning pruning and its corresponding thinning ratio.

[0082] It should be noted that the pruning plan in this application refers to an executable pruning instruction that requires the removal of leaves or branches in the target area. The purpose of determining the pruning plan is to transform the spatial location information of canopy water imbalance into a quantitatively executable pruning decision. By clarifying the specific location where pruning is required and the proportion of branches and leaves to be removed from the branch group in that location, the leaf area and transpiration demand of the canopy branch group in the priority stress location are selectively reduced, so that the water consumption in that location is reduced to a level that matches the local water supply capacity of the soil. This alleviates the problem of uneven water distribution within the canopy, ensures that the intensity of pruning intervention is proportional to the degree of water imbalance in each location, and maximizes the preservation of the photosynthetic production potential of the branch group in the non-stress location while eliminating local water stress, thus maintaining the overall yield base of the canopy after pruning.

[0083] In step S104, after the thinning and pruning is completed according to the thinning and pruning scheme, the water compensation priority coefficient, which characterizes the soil's effective water compensation capacity for the canopy structure after pruning, is calculated from the global recovery slope of the fruit tree's full canopy transpiration rate change curve during a preset period after pruning and the global attenuation slope during the stress period before pruning.

[0084] In some embodiments, the global recovery slope of the fruit tree canopy transpiration rate change curve within a preset time period after pruning is determined by the following steps:

[0085] After the pruning was completed, the transpiration rate change curves of leaves in all directions of the fruit tree canopy were collected within a preset time period. The transpiration rate change curves of leaves in all directions were weighted and averaged according to the same time stamp to obtain the first transpiration rate change curve of the entire fruit tree canopy.

[0086] Linear regression fitting is performed on the first transpiration rate change curve, and the slope of the fitted straight line is extracted as the global recovery slope of the fruit tree full canopy transpiration rate change curve within a preset time period after pruning.

[0087] In practice, firstly, within a predetermined time period after the completion of pruning, which is a fixed observation duration set according to the physiological cycle of water stress recovery of fruit tree varieties, for example, a continuous 48 hours after pruning, the transpiration rate change curves of leaves in each direction are simultaneously collected using thermal infrared imagers fixed in each direction. Using the timestamp of each sampling point in each curve as the alignment reference, the transpiration rate values ​​of different directions at the same timestamp are summed with equal weights, that is, each direction is assigned the same weight coefficient of 0.25. After processing the entire predetermined time period by timestamp, the first transpiration rate change of the entire canopy of the fruit tree is obtained. The first transpiration rate change curve refers to a synthetic data sequence that characterizes the temporal variation of the overall transpiration rate of the fruit tree canopy within a preset time period after the completion of thinning and pruning. Then, all time points and corresponding transpiration rate values ​​in the first transpiration rate change curve are used as sample point pairs, and the least squares linear regression method is used to fit a straight line. The least squares linear regression method solves for the optimal slope and intercept parameters by minimizing the sum of squares of the residuals between the observed transpiration rate values ​​of each sample point and the predicted values ​​of the fitted straight line. The values ​​of the coefficients of the first-order terms in the fitted straight line equation are extracted as the global recovery slope of the transpiration rate change curve of the entire fruit tree canopy within the preset time period after pruning.

[0088] It should be noted that, in this application, the global recovery slope refers to the average growth rate of the overall transpiration rate of the canopy after pruning over time. Since pruning reduces part of the canopy leaf area, the transpiration rate of the remaining canopy branches will change dynamically due to the redistribution of soil moisture supply. The rate of increase of this change in the time series directly reflects the compensatory water supply capacity of the effective soil moisture to the remaining branches after pruning. That is, whether the remaining branches and leaves can obtain sufficient soil moisture to restore and enhance transpiration activity after pruning relieves the water competition of the stressed leaves. Therefore, determining the global recovery slope is to transform the recovery process of the canopy transpiration capacity after pruning into a quantifiable rate index, providing a data basis for subsequent ratio calculations with the global attenuation slope during the pre-pruning stress period. Then, a water compensation priority coefficient is constructed through the slope recovery ratio to numerically assess the degree of compensation of the effective soil moisture to the canopy structure after pruning and determine whether supplementary irrigation needs to be triggered.

[0089] In some embodiments, the global attenuation slope during the pre-pruning stress period is determined using the following steps:

[0090] The transpiration rate change curves of leaves in all directions were obtained from the initial moment of local water stress to the start of thinning pruning. The transpiration rate change curves of leaves in all directions were weighted and averaged according to the same time stamp to obtain the second transpiration rate change curve of the entire canopy of the fruit tree.

[0091] Linear regression was performed on the second transpiration rate change curve, and the slope of the fitted line was extracted as the global decay slope during the stress period before pruning.

[0092] In practice, firstly, the initial moment of the first occurrence of local water stress is taken as the starting point of the time period, and the moment when thinning pruning begins is taken as the ending point. The transpiration rate change curves of leaves in each direction are extracted from the stored dataset of transpiration rate change curves in each direction. The transpiration rate values ​​of each direction at the same time stamp are then summed using an equal-weighted arithmetic average, i.e., each direction is assigned the same weight coefficient of 0.25. After processing the entire stress period time stamp by time stamp, the second transpiration rate change curve of the entire fruit tree canopy is obtained. The second transpiration rate change curve of the entire fruit tree canopy refers to... A synthetic data sequence of the overall canopy transpiration rate changes over time from the onset of water stress to before pruning intervention was generated. Then, all time points and corresponding transpiration rate values ​​in the second transpiration rate change curve of the entire canopy were used as sample point pairs. The least squares linear regression method was used to fit a straight line. The least squares linear regression method solves for the optimal slope and intercept parameters by minimizing the sum of squares of the residuals between the observed transpiration rate values ​​of each sample point and the predicted values ​​of the fitted straight line. The values ​​of the coefficients of the first-order terms in the fitted straight line equation were extracted as the global attenuation slope during the stress period before pruning.

[0093] It should be noted that, in this application, the global attenuation slope refers to the average rate of decrease in the overall transpiration rate of the canopy before pruning during the duration of water stress. The purpose of determining the global attenuation slope is to quantify the entire process of water stress from its occurrence to its accumulation before pruning into a calculable rate of change index, providing a numerical basis for assessing the severity of stress. The larger the absolute value of the global attenuation slope, the faster the overall transpiration rate of the canopy decreases during the stress period and the higher the stress intensity. Conversely, the smaller the value, the slower the stress development. In subsequent steps, the global attenuation slope is used as the denominator to perform a ratio calculation with the global recovery slope after pruning to construct the slope recovery ratio. This allows the calculation of the water compensation priority coefficient to be based on a comparison of the dynamic changes in the transpiration rate of the same canopy before and after pruning, thereby accurately reflecting the actual compensating capacity of the effective soil water for stress damage after pruning intervention.

[0094] In some embodiments, after thinning pruning is completed according to the thinning pruning scheme, the water compensation priority coefficient, which characterizes the soil's effective water compensation capacity for the canopy structure after pruning, is calculated from the global recovery slope of the fruit tree's full canopy transpiration rate change curve during a preset period after pruning and the global attenuation slope during the stress period before pruning, using the following steps:

[0095] The slope recovery ratio is obtained by taking the absolute value of the global recovery slope of the fruit tree's full canopy transpiration rate change curve during the preset period after pruning and the absolute value of the global decay slope during the stress period before pruning.

[0096] The slope recovery ratio is input into a preset S-shaped function normalization model, and the output is a water compensation priority coefficient that characterizes the soil's effective water compensation capacity for the canopy structure after pruning.

[0097] In practice, firstly, the absolute value of the global recovery slope of the determined fruit tree canopy transpiration rate change curve within a preset time period after pruning is taken, and the absolute value of the global decay slope within the determined stress period before pruning is also taken. The global recovery slope is positive, and the global decay slope is negative; both are positive real numbers after taking their absolute values. A division operation is performed using the absolute value of the global recovery slope as the numerator and the absolute value of the global decay slope as the denominator. The resulting quotient is the slope recovery ratio. The slope recovery ratio refers to the ratio of the recovery rate of the canopy transpiration rate after pruning to the decay rate of the transpiration rate during the stress period before pruning. A ratio greater than 1 indicates that the recovery rate exceeds the decay rate, while a ratio between 0 and 1 indicates that the recovery rate is less than the decay rate. Next, the slope recovery ratio is input into a preset S-shaped function normalization model. The model employs a nonlinear mapping transformation. The S-shaped function normalization model adopts the standard form of the Logistic function, specifically expressed as: the water compensation priority coefficient equals 1 divided by 1 plus the negative k of the natural exponential function e multiplied by the difference between the slope recovery ratio and x0 raised to the power of 1. Here, k is the shape parameter controlling the steepness of the curve, and x0 is the position parameter of the curve center offset. k and x0 are pre-calibrated using controlled stress recovery experiments based on historical measured data. For example, k is calibrated to 3 and x0 to 1, so that when the slope recovery ratio is 0, the model output approaches 0; when the slope recovery ratio is equal to x0, the output is 0.5; and when the slope recovery ratio increases, the output asymptotically approaches 1. This compresses and maps the slope recovery ratio from the positive domain to the interval between 0 and 1, outputting the water compensation priority coefficient that characterizes the soil's effective water compensation capacity for the pruned canopy structure.

[0098] It should be noted that the water compensation priority coefficient in this application refers to the priority level index of the soil's effective water compensation capacity for the pruned canopy structure. The closer the value is to 1, the more sufficient the soil's effective water supply capacity for the remaining canopy after pruning is, and the lower the priority of irrigation compensation. Unlike the passive response mode of existing technologies that generally rely on a single soil moisture content threshold or a fixed schedule to trigger irrigation, the water compensation priority coefficient constructs a two-way comparison mechanism of the dynamic changes in the canopy transpiration rate before and after pruning in its determination logic. It takes the ratio of the global recovery slope after pruning to the global decay slope before pruning as input, and uses an S-shaped function to nonlinearly map the ratio of the recovery rate to the decay rate into a continuous priority level output. This coefficient integrates the two dimensions of the physical effect of canopy pruning intervention and the physiological response of soil moisture, so that irrigation decisions are no longer based on the absolute value of soil moisture in isolation, but dynamically assess the actual compensation efficiency of soil moisture for the remaining branches after canopy structure regulation. It couples the two management links of pruning and irrigation into an integrated parameter for collaborative decision-making through the time-series response of transpiration rate.

[0099] In step S105, when the water compensation priority coefficient is less than the preset priority threshold, irrigation control is performed according to the pre-set irrigation plan during the plant's water self-buffering period until the water compensation priority coefficient rises back to the preset priority threshold, at which point the normal irrigation decision is restored.

[0100] In some embodiments, when the water compensation priority coefficient is less than a preset priority threshold, irrigation control is performed according to a pre-set irrigation plan within the plant's water self-buffering period until the water compensation priority coefficient rises back to the preset priority threshold. The resumption of normal irrigation decisions is achieved through the following steps:

[0101] The water compensation priority coefficient is compared with the preset priority threshold. If the water compensation priority coefficient is less than the preset priority threshold, the plant water self-buffering irrigation mode is triggered and the preset irrigation scheme is started. The preset irrigation scheme is to continuously irrigate the pruned canopy with a preset high-frequency small quota irrigation cycle.

[0102] During the execution of the plant water self-buffered irrigation mode, thermal infrared image sequences of leaf surfaces in all directions are re-acquired at preset time intervals and the water compensation priority coefficient is updated. The updated water compensation priority coefficient is continuously compared with the preset priority threshold.

[0103] When the updated water compensation priority coefficient rises to a level greater than or equal to the preset priority threshold, the plant water self-buffering irrigation mode is terminated, and the regular irrigation decision is restored.

[0104] In practice, firstly, the water compensation priority coefficient is compared with a preset priority threshold. The preset priority threshold is a critical value pre-calibrated based on the minimum acceptable level of water compensation efficiency for fruit tree varieties, for example, set to 0.5. When the water compensation priority coefficient is less than 0.5, it indicates that the effective soil moisture after pruning is insufficient to compensate for the water supply to the remaining canopy branches, and the recovery rate of canopy transpiration rate is significantly lower than the 50% mapping level of the attenuation rate during the stress period. At this time, the plant's self-buffered irrigation mode is triggered. The plant's self-buffered irrigation mode refers to the irrigation mode when the canopy has not fully recovered after pruning. When the water balance is restored, a high-frequency, low-quota compensatory irrigation operation is adopted. Simultaneously, a pre-set irrigation plan under this mode is activated, which involves continuous irrigation of the pruned canopy using a preset high-frequency, low-quota irrigation cycle. This preset high-frequency, low-quota irrigation cycle consists of two parameters: single irrigation depth and irrigation interval. The single irrigation depth is set to 30% to 50% of the conventional irrigation single irrigation volume, and the irrigation interval is set to half of the conventional irrigation interval. For example, if conventional irrigation involves 20 mm of water per irrigation at 48-hour intervals, then the high-frequency, low-quota irrigation plan involves 8 mm of water per irrigation at 48-hour intervals. After a 24-hour interval, during the self-buffered irrigation mode, a preset time interval is used as the update cycle. The preset time interval is set according to the detectable rate of change in canopy transpiration response within the irrigation cycle, for example, 6 hours. Every 6 hours, a complete time-series thermal infrared image of the leaf surface in each direction is re-acquired using thermal infrared imagers and micro-weather stations. The water compensation priority coefficient is recalculated according to the complete process of extracting the transpiration rate change curve, determining the relief of local water stress, updating and calculating the global recovery slope and global attenuation slope, and updating the water compensation priority coefficient. The updated water compensation priority coefficient is continuously compared with the preset priority threshold of 0.5. Continuous comparison means that a threshold comparison is performed immediately after each update cycle and the comparison result is recorded. Finally, when the updated water compensation priority coefficient rises to greater than or equal to 0.5, it indicates that the soil's effective water compensation capacity for the pruned canopy structure has reached an acceptable level. At this time, the self-buffered irrigation mode is terminated, and the irrigation controller is switched back to the conventional irrigation decision mode based on soil moisture content or evapotranspiration feedback, restoring conventional irrigation decisions.

[0105] It should be noted that, in this application, routine irrigation decision refers to the daily irrigation system established before the occurrence of local water stress and before the implementation of thinning and pruning, including the decision logic of irrigation time and amount based on evapotranspiration forecast or soil moisture threshold, which will not be elaborated here.

[0106] Furthermore, in another aspect of this application, in some embodiments, this application provides an intelligent pruning and water management collaborative control system for fruit tree cultivation, referencing... Figure 4The figure is a schematic diagram of the structure of an intelligent pruning and water management collaborative control system for fruit tree cultivation according to some embodiments of this application. The intelligent pruning and water management collaborative control system for fruit tree cultivation includes: an acquisition module 201, a processing module 202, and an execution module 203, which are described below:

[0107] The acquisition module 201 in this application is mainly used to acquire thermal infrared image sequences of the leaf surface in different directions of the fruit tree canopy and extract the transpiration rate change curves of the leaves in each direction.

[0108] Processing module 202, in this application, is mainly used to identify the initial moment of local water stress of the corresponding leaf based on the transpiration rate change curve of the leaf in each direction.

[0109] The processing module 202 is also used to construct a thinning and pruning scheme based on the orientation of the first occurrence of local water stress and the transpiration contribution of the fruit tree canopy branch group when the time difference between the initial moments of local water stress between any two directional leaves reaches a preset stress threshold.

[0110] In addition, the processing module 202 is also used to calculate, after the pruning is completed according to the pruning plan, a water compensation priority coefficient that characterizes the soil's effective water compensation capacity for the pruning canopy structure by the global recovery slope of the fruit tree's full canopy transpiration rate change curve during a preset period after pruning and the global attenuation slope during the stress period before pruning.

[0111] The execution module 203 in this application is mainly used to execute irrigation control according to the pre-set irrigation plan during the plant water self-buffering period when the water compensation priority coefficient is less than the preset priority threshold, until the water compensation priority coefficient rises back to the preset priority threshold, and then resumes the normal irrigation decision.

[0112] In addition, this application also provides a computer device, which includes a memory and a processor. The memory stores code, and the processor is configured to acquire the code and execute the above-described intelligent pruning and water management collaborative control method for fruit tree cultivation.

[0113] In some embodiments, reference Figure 5 The figure is a schematic diagram of the structure of a computer device for implementing a method for intelligent pruning and water management in fruit tree cultivation, according to some embodiments of this application. The method for intelligent pruning and water management in fruit tree cultivation described in the above embodiments can... Figure 5 The computer device shown is used to implement this, and the computer device includes at least one processor 301, a communication bus 302, a memory 303, and at least one communication interface 304.

[0114] The processor 301 can be a general-purpose central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more devices used to control the execution of the intelligent pruning and water management collaborative control method for fruit tree cultivation in this application.

[0115] The communication bus 302 can be used to transmit information between the aforementioned components.

[0116] The memory 303 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disks or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory 303 may exist independently and be connected to the processor 301 via the communication bus 302. The memory 303 may also be integrated with the processor 301.

[0117] The memory 303 stores program code for executing the scheme of this application, and its execution is controlled by the processor 301. The processor 301 executes the program code stored in the memory 303. The program code may include one or more software modules. In the above embodiments, the determination of the intelligent pruning and water management collaborative control method for fruit tree cultivation can be achieved by the processor 301 and one or more software modules in the program code in the memory 303.

[0118] Communication interface 304 uses any transceiver-like device for communicating with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0119] In a specific implementation, as one example, a computer device may include multiple processors, each of which may be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. Here, a processor may refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0120] The aforementioned computer device can be a general-purpose computer device or a special-purpose computer device. In specific implementations, the computer device can be a desktop computer, a portable computer, a network server, a handheld digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. This application does not limit the type of computer device.

[0121] In addition, this application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned intelligent pruning and water management collaborative control method for fruit tree cultivation.

[0122] Although preferred embodiments of this application have been described, those skilled in the art, once they have learned the basic inventive concept, can make other changes and modifications to these embodiments.

[0123] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application.

Claims

1. A method for intelligent pruning and water management coordinated control in fruit tree cultivation, characterized in that, Includes the following steps: Thermo-infrared image sequences of leaf surfaces at different locations in the canopy of fruit trees were obtained, and the transpiration rate variation curves of leaves at each location were extracted. The initial moment of local water stress for leaves in each direction was identified based on the transpiration rate change curves of the leaves in each direction. When the time difference between the initial moments of local water stress between any two directional leaves reaches the preset stress threshold, a thinning and pruning scheme is constructed based on the orientation of the initial moment of local water stress and the transpiration contribution of the fruit tree canopy branch group. After pruning according to the aforementioned pruning plan, the transpiration rate change curves of leaves at various locations within a preset time period after pruning are collected. These curves are then weighted and averaged according to the same time stamp to obtain the first transpiration rate change curve of the entire canopy. A linear regression is performed on this first transpiration rate change curve, and the slope of the fitted line is extracted as the global recovery slope of the transpiration rate change curve of the entire canopy within the preset time period after pruning. The global recovery slope refers to the average growth rate of the overall transpiration rate of the canopy after pruning over time. The transpiration rate change curves of leaves at various locations are then obtained from the initial moment of local water stress to the start of pruning. The rate change curves are weighted and averaged according to the same time stamp to obtain the second transpiration rate change curve of the entire canopy of the fruit tree; linear regression is performed on the second transpiration rate change curve, and the slope of the fitted line is extracted as the global decay slope during the stress period before pruning. The global decay slope refers to the average rate of decrease of the overall canopy transpiration rate before pruning during the water stress period; the absolute value of the global recovery slope of the fruit tree's entire canopy transpiration rate change curve during the preset period after pruning is taken as the ratio to the absolute value of the global decay slope during the stress period before pruning, to obtain the slope recovery ratio; the slope recovery ratio is input into the preset S-shaped function normalization model, and the water compensation priority coefficient, which characterizes the soil's effective water compensation capacity for the canopy structure after pruning, is output; When the water compensation priority coefficient is less than the preset priority threshold, irrigation control is performed according to the pre-set irrigation plan during the plant's water self-buffering period until the water compensation priority coefficient rises back to the preset priority threshold, at which point the normal irrigation decision is restored.

2. The method as described in claim 1, characterized in that, Acquire thermal infrared image sequences of leaf surfaces at different locations within the fruit tree canopy, and extract transpiration rate variation curves for each location. Specifically, this includes: A thermal infrared image sequence of the leaf surface at different positions of the fruit tree canopy was acquired using a thermal infrared imager, and the canopy orientation was marked and time stamped in the thermal infrared image sequence. Based on the thermal infrared image sequence after aligning the canopy orientation markers with the timestamps, pixel-level continuous temperature sequences of leaves in each orientation are extracted by segmenting the leaf area and the background area by setting grayscale thresholds. Based on the pixel-level continuous temperature sequence of leaves in each direction and the ambient air temperature, relative humidity and net radiation flux of the canopy collected by the micro-weather station, the temporal stomatal conductance of leaves in each direction is inverted based on the canopy energy balance equation. The transpiration rate variation curves of the blades in each direction were constructed based on the temporal stomatal conductance of the blades in each direction.

3. The method as described in claim 1, characterized in that, Based on the transpiration rate variation curves of leaves in each direction, the initial moment of local water stress for leaves in the corresponding directions was identified, specifically including: The transpiration rate variation curves of the leaves in each direction are filtered and smoothed to obtain the noise-reduced transpiration rate variation curves of the leaves in each direction. The first-order difference operation is performed on the denoised transpiration rate variation curves of the blades in each direction to obtain the first-order difference sequence of transpiration rate of the blades in each direction. Extract the first inflection point from the first-order difference sequence of transpiration rate of leaves in each direction, where the value changes from positive to negative and remains negative for the next three consecutive sampling points. Mark the time corresponding to this inflection point as the initial moment of local water stress for the corresponding leaf.

4. The method as described in claim 1, characterized in that, The transpiration contribution of the fruit tree canopy branch groups was determined using the following steps: The transpiration rate variation curves of leaves in each position of the fruit tree canopy before thinning and pruning are obtained. The transpiration rate variation curves of leaves in each position are integrated in the time domain to obtain the transpiration rate per unit area of ​​leaves in each position. Based on the transpiration rate per unit area of ​​leaves in each direction and the leaf area index of the corresponding canopy branch group, the total transpiration water consumption of the canopy branch group in each direction is calculated. The transpiration contribution of each canopy branch is obtained by dividing the total transpiration water consumption of each canopy branch by the sum of the total transpiration water consumption of all canopy branches.

5. The method as described in claim 1, characterized in that, When the time difference between the initial moments of local water stress between any two directional leaves reaches a preset stress threshold, a thinning and pruning plan is constructed based on the location of the earliest initial moment of local water stress and the transpiration contribution of the fruit tree canopy branches. This plan specifically includes: Compare the initial time of local water stress in any two directions of the leaf, calculate the time difference between the initial time of local water stress between each pair of directions, and when the time difference between the initial time of local water stress between any pair of leaves reaches the preset stress threshold, mark the direction where the initial time of local water stress occurs first as the priority stress direction. The transpiration contribution of the canopy branch group in the priority stress direction is compared with the preset contribution threshold. If the transpiration contribution of the canopy branch group in the priority stress direction exceeds the preset contribution threshold, then the priority stress direction is determined as the target direction for thinning and pruning. Based on the difference between the transpiration contribution of the canopy branch group in the target pruning direction and the average transpiration contribution of the canopy branch groups in all directions, the pruning amount ratio in the target pruning direction is determined according to a direct proportional relationship, and a pruning plan containing the pruning target pruning direction and the corresponding pruning amount ratio is generated.

6. The method as described in claim 1, characterized in that, The different orientations of the fruit tree canopy include due east, due west, due south, and due north, with the center of the fruit tree trunk as the orthogonal reference.

7. A smart pruning and water management collaborative control system for fruit tree cultivation, used to execute the smart pruning and water management collaborative control method for fruit tree cultivation as described in any one of claims 1 to 6, characterized in that, The system includes: The acquisition module is used to acquire thermal infrared image sequences of the leaf surface at different positions in the canopy of fruit trees and extract the transpiration rate change curves of the leaves at each position. The processing module is used to identify the initial moment of local water stress for leaves in each direction based on the transpiration rate change curves of the leaves in each direction. The processing module is also used to construct a thinning and pruning scheme based on the orientation of the first occurrence of local water stress and the transpiration contribution of the fruit tree canopy branch group when the time difference between the initial moments of local water stress between any two directional leaves reaches a preset stress threshold. The processing module is also used to calculate, after the pruning is completed according to the pruning plan, a water compensation priority coefficient that characterizes the soil's effective water compensation capacity for the pruning canopy structure by using the global recovery slope of the fruit tree's full canopy transpiration rate change curve during a preset period after pruning and the global attenuation slope during the stress period before pruning. The execution module is used to perform irrigation control according to the pre-set irrigation plan during the plant's water self-buffering period when the water compensation priority coefficient is less than the preset priority threshold, until the water compensation priority coefficient rises back to the preset priority threshold, and then resumes the normal irrigation decision.

8. A computer device, characterized in that, The computer device includes a memory and a processor, the memory storing code, and the processor being configured to retrieve the code and execute the intelligent pruning and water management collaborative control method for fruit tree cultivation as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the intelligent pruning and water management collaborative control method for fruit tree cultivation as described in any one of claims 1 to 6.

Images