CN-122022286-A - Pear ecological index evaluation method based on comprehensive data analysis

CN122022286ACN 122022286 ACN122022286 ACN 122022286ACN-122022286-A

Abstract

The invention discloses a pear tree ecological index evaluation method based on comprehensive data analysis, and relates to the technical field of intelligent agriculture. The method comprises the steps of obtaining tree physiological data, environment stress data and resource input data in real time, dividing a pear physiological system into a fruit growth subsystem, a root system development subsystem, an insect disease resistance defense subsystem and a symbiotic support subsystem, constructing a resource allocation game optimization model by taking all the subsystems as game participants, dynamically calculating the current capital state and environment stress index of all the game participants based on the collected multidimensional data, solving by adopting a reinforcement learning algorithm to obtain an optimal resource allocation vector by taking the pear total resource budget as a constraint condition, calculating pear ecological harmony index by comparing the optimal resource allocation vector with the actually measured allocation vector, and generating and outputting a corresponding ecological management strategy based on the calculated ecological harmony index, the optimal resource allocation vector and the current capital state of all the game participants.

Inventors

XU LEI
XIONG CHUANYONG
ZHOU CHAOHUA
YANG SICHAO
WANG YUN
LI YANTING

Assignees

江西省农业科学院园艺研究所

Dates

Publication Date: 20260512
Application Date: 20260115

Claims (10)

1. The pear ecological index evaluation method based on comprehensive data analysis is characterized by comprising the following steps of: Establishing a pear multi-dimensional data acquisition system, and acquiring tree physiological data, environmental stress data and resource input data in real time; Dividing the pear physiological system into fruit growth subsystems Root system development subsystem Disease and insect resistant defense subsystem And symbiotic support subsystem Constructing a resource allocation game optimization model by taking each subsystem as a game participant; Dynamically calculating a current capital status of each of the gaming participants based on the collected multidimensional data And an environmental stress index ; With pear tree total resource budget ; Taking pear total resource budget as constraint condition, solving the resource allocation game optimization model by adopting reinforcement learning algorithm to obtain an optimal resource allocation vector ; By comparing optimal resource allocation vectors And the actually measured distribution vector Calculating pear ecological harmony index ; Based on the ecological coordination index Optimal resource allocation vector And the current capital status of each of said game participants And generating and outputting a corresponding ecological management strategy.
2. The pear ecology index evaluation method based on comprehensive data analysis of claim 1, wherein the construction flow of the resource allocation game optimization model comprises the following steps: defining the resource allocation proportion of each subsystem in each time period t, wherein the sum of the proportions is 1; Establishing a benefit function of the ith subsystem at stage t ; Constructing the resource allocation game optimization model by taking the discount maximization of the sum of the benefits of all subsystems in the growth stage T as an optimization target, wherein the optimization target is defined as: wherein As a time-discounting factor, the time-critical factor, Is the time-varying weight coefficient of the ith subsystem.
3. The pear ecology index evaluation method based on comprehensive data analysis of claim 2, wherein the profit function Depending on the amount of resources allocated, the resource allocation of other subsystems, and the environmental stress level, at least: based on the logarithmic benefit of the allocated resources, the benefit decreases as the resources increase; a competition loss term with other functional subsystems, wherein the competition loss is determined by a competition strength matrix; an environmental stress penalty term, and triggering a penalty when the functional subsystem capital is below a maintenance threshold; The benefit function The definition is as follows: ; Wherein, the For the phase-related priority coefficient, As a coefficient of the efficiency of the utilization of the resources, In order to compete for the sensitivity coefficient, In order to compete for the strength matrix elements, For the environmental stress penalty factor, A resource threshold is maintained for each participant's function, To take 0 and Is operated by the maximum value of (a).
4. A pear ecology index evaluation method based on comprehensive data analysis as described in claim 3 wherein the competitive strength matrix element Dynamically setting according to the cooperative or competitive relation among all subsystems, and dynamically adjusting according to the following rules: , wherein, For the reference contention strength, k is the reference contention strength, Scaling factors for capital status; In particular: setting according to the cooperative relationship between the fruit growing subsystem and the symbiotic supporting subsystem ; Setting according to the strong competition relationship between the root system development subsystem and the disease and insect resistance defense subsystem 。
5. The pear ecology index evaluation method based on comprehensive data analysis of claim 2, wherein the time-varying weight coefficient Dynamic adjustment is carried out according to the degree of environmental stress, and the adjustment rule is as follows: ; Wherein, the For the base weight of the ith game participant, And In order for the stress response parameter to be a function of, In order to indicate the function, A stress response threshold for the ith game participant.
6. The pear ecology index evaluation method based on comprehensive data analysis of claim 1, wherein the solving the resource allocation game optimization model using a reinforcement learning algorithm comprises: An upper central coordinator taking as input the capital status, environmental stress index, climate period and total resources of each subsystem, its status space Encoding the weathered period by using the resource allocation vector As output, the rewarding function of the upper central coordinator Defining by comprehensively considering the current benefit, future benefit expectations and stability of each capital state of each functional subsystem; A lower subsystem policy network that takes as input the capital status, environmental stress, last-stage allocation, and profit of each subsystem, and outputs a desired resource proportion; The reinforcement learning algorithm iteratively updates parameters of the upper central coordinator and the lower subsystem strategy network in a fixed strategy alternating optimization mode; The rewarding function of the upper central coordinator The definition is as follows: ; Wherein, the For the future benefits the desired weight is given, For the capital fluctuation penalty factor, As a function of the standard deviation of the signal, The cost function is a predicted value representing the cumulative expected benefit obtained by the system from the current time step to a future period.
7. The method for evaluating pear ecology index based on comprehensive data analysis according to claim 1, wherein the pear ecology harmony index The calculation is performed by the following formula: , wherein, Representing Euclidean norms, the measured allocation vector Obtained by a carbon isotope tracing method or a back-pushing method based on the growth of each organ; In particular: When (when) And if the ecological coordination is determined to be unbalanced, executing the generation and outputting the corresponding ecological management strategy.
8. The pear ecology index evaluation method based on comprehensive data analysis of claim 1, wherein the generating of the ecology management policy comprises: calculating the allocation bias of the ith game participant ; Matching a preset prescription rule according to the distribution deviation; The prescriptions are individually adjusted and quantified by combining the ages, variety characteristics and the feasibility of agriculture; Converting the final prescription into an augmented reality guiding instruction to guide field operation; wherein the prescription rule at least comprises: Rule 1 when And is also provided with In this case, a root-promoting management prescription is generated, A target capital status for the 2 nd game participant; Rule 2 when And when the disease index in D (t) is more than 0.6, generating a defense-enhanced prescription; Rule 3 when For 3 phases <0.1, a symbiotic system repair prescription is generated.
9. A pear ecology index evaluation system based on comprehensive data analysis, wherein the system is used for realizing the pear ecology index evaluation method based on comprehensive data analysis as claimed in any one of claims 1 to 8, comprising: The multi-source data acquisition module is used for acquiring physiological, environmental stress and resource input data of the tree body in real time; A gaming modeling and solving module comprising: the resource allocation game optimization model construction unit is used for constructing a resource allocation game optimization model with the discount maximization of the sum of the benefits of all subsystems in the growth stage T as an optimization target; For solving a resource allocation game optimization model using reinforcement learning algorithms to obtain optimal resource allocation vectors A hierarchical reinforcement learning solving unit of (1); the ecological index calculation module is used for comparing the optimal resource allocation vectors And the actually measured distribution vector Calculating pear ecological harmony index ; A management prescription generation module for generating a management prescription based on the ecological coordination index Optimal resource allocation vector And the current capital status of each game participant Generating and outputting a corresponding ecological management strategy; And the visual interaction module is used for displaying the evaluation result of the pear ecological index to the user and providing augmented reality operation guidance.
10. A computer readable storage medium comprising a memory and a processor, the memory having stored therein a degree of computers and the computer program when executed by the processor implementing the pear tree ecological index assessment method based on resource allocation gaming and dynamic optimization of any of claims 1-8.

Description

Pear ecological index evaluation method based on comprehensive data analysis Technical Field The invention relates to the technical field of intelligent agriculture, in particular to a pear tree ecological index evaluation method based on comprehensive data analysis. Background The pear tree is used as one of main economic fruit trees, and ecological and sustainable cultivation of the pear tree is a key for guaranteeing fruit safety and improving economic benefit and ecological benefit of an orchard. Traditional pear garden management is mostly dependent on farmer experience or single agronomic index (such as yield, leaf nutrition and the like), and lacks accurate quantitative evaluation on dynamic processes and multi-objective coordination in pear ecological system. At present, for ecological evaluation and management of pear trees, the prior art mainly focuses on the relationship between external input (such as water and fertilizer) and final output (if actual output), but the dynamic allocation mechanism of the photosynthetic products between different functional organs (such as fruits, leaves, root systems and defensive tissues) in the tree body lacks effective monitoring and regulating means, so that operable dynamic regulating suggestions based on real-time data cannot be provided. In addition, although technologies such as the internet of things and sensors are applied to orchard information acquisition, mass data are generated for state monitoring and simple early warning, and the internal game process and regulation and control rules of the ecosystem reflected after the data are not deeply mined. The existing model is concentrated on growth prediction or disease early warning, and rarely starts from the whole ecological system, simulates dynamic games which are carried out by all functional subsystems in the tree body for competing for limited resources, and generates research of an optimization management strategy according to the dynamic games. Therefore, a pear ecological index evaluation method based on comprehensive data analysis is provided. Disclosure of Invention The invention mainly aims to provide a pear ecological index assessment method based on comprehensive data analysis, which can be used for quantifying the internal resource competition and cooperation of each subsystem of a pear and optimizing multi-objective long-term balance by constructing a digital twin system with a resource allocation game optimizer as a core, realizing accurate management based on data driving and effectively solving the problems in the background technology. In order to achieve the above purpose, the invention adopts the technical proposal that, The pear ecological index evaluation method based on comprehensive data analysis comprises the following steps: Establishing a pear multi-dimensional data acquisition system, and acquiring tree physiological data, environmental stress data and resource input data in real time; Dividing the pear physiological system into fruit growth subsystems Root system development subsystemDisease and insect resistant defense subsystemAnd symbiotic support subsystemConstructing a resource allocation game optimization model by taking each subsystem as a game participant; Dynamically calculating a current capital status of each of the gaming participants based on the collected multidimensional data And an environmental stress index; With pear tree total resource budgetSolving the resource allocation game optimization model by adopting a reinforcement learning algorithm as constraint conditions to obtain an optimal resource allocation vector; By comparing optimal resource allocation vectorsAnd the actually measured distribution vectorCalculating pear ecological harmony index; Based on the ecological coordination indexOptimal resource allocation vectorAnd the current capital status of each of said game participantsAnd generating and outputting a corresponding ecological management strategy. Further, the construction flow of the resource allocation game optimization model comprises the following steps: defining the resource allocation proportion of each subsystem in each time period t, wherein the sum of the proportions is 1; Establishing a benefit function of the ith subsystem at stage t ; Constructing the resource allocation game optimization model by taking the discount maximization of the sum of the benefits of all subsystems in the growth stage T as an optimization target, wherein the optimization target is defined as: wherein As a time-discounting factor, the time-critical factor,Is the time-varying weight coefficient of the ith subsystem. Further, the benefit functionDepending on the amount of resources allocated, the resource allocation of other subsystems, and the environmental stress level, at least: based on the logarithmic benefit of the allocated resources, the benefit decreases as the resources increase; a competition loss term with other functional subsystems, wherein t