Insect materialization control gradient decision-making method and system based on multi-mode perception driving

Abstract

The invention relates to a pest physicochemical control gradient decision-making method and system based on multi-mode perception driving, and belongs to the field of agricultural pest control. The method comprises the steps that multi-modal perception data are acquired, a multi-modal perception model is constructed, perception information is output, and the perception information comprises insect quantity density, a voiceprint migration index and an environmental stress coefficient; constructing a cellular automaton model based on the perception information, simulating pest diffusion and outputting a risk thermodynamic diagram; and outputting a pest physicochemical control gradient decision based on the risk thermodynamic diagram. The method solves the long-standing problems of response lag, too high cost, inaccurate prediction and the like in the field of pest control, and has an application prospect.

Description

Technical Field

[0001] The present invention belongs to the technical field of agricultural pest control, and specifically relates to a gradient decision-making method and system for pest physical and chemical control based on multimodal perception. Background Art

[0002] Globally, pests cause severe crop yield losses annually, resulting in direct economic losses exceeding $100 billion. Traditional pest control faces three major challenges: 1) Pest population dynamics are spatiotemporally heterogeneous, making it difficult to accurately assess risk levels using a single monitoring method; 2) Overuse of chemical pesticides has led to increased resistance; and 3) Existing decision-making systems often rely on static thresholds and are unable to dynamically respond to the nonlinear coupling effects of environmental stress and pest migration. Regarding perception technology, visual sensors (such as spectral imaging) can detect insect population density but cannot characterize pest migration behavior (such as flight and aggregation). Soundprint monitoring technology has been applied to social insects such as termites, but lacks a characteristic frequency weighting model and results in high false alarm rates. Environmental parameter monitoring often uses temperature and humidity thresholds, failing to establish a dynamic correlation between environmental stress and pest physiological responses. Regarding predictive models, traditional diffusion models (such as the logistic model) ignore spatial heterogeneity and result in high prediction errors. Geographic Information Systems (GIS) can generate risk maps but lack the spatiotemporal evolution mechanisms of cellular automata. Machine learning models rely on historical data and are slow to respond to sudden pest outbreaks. Therefore, a multimodal perception-driven gradient decision-making method for pest physical and chemical control is urgently needed to solve the above problems. Summary of the Invention

[0003] In order to solve the above problems existing in the prior art, the present invention provides a pest physical and chemical control gradient decision-making method and system based on multimodal perception drive.

[0004] The purpose of the present invention can be achieved through the following technical solutions:

[0005] A multimodal perception-driven gradient decision-making method for pest physical and chemical control, wherein the implementation of the gradient decision-making method for pest physical and chemical control comprises the following steps:

[0006] S1: Acquire multimodal perception data and build a multimodal perception model to output perception information, including insect population density, soundprint migration index, and environmental stress coefficient;

[0007] S2: constructing a cellular automation model based on the perception information, simulating the spread of pests and outputting a risk heat map;

[0008] S3: Outputting a gradient decision on the physical and chemical pest control based on the risk heat map.

[0009] Preferably, the step S1 specifically includes:

[0010] S101: Outputting insect population density and insect population change rate through visual sensors;

[0011] S102: Capture the sounds of pest activity, extract characteristic frequencies, and define a soundprint migration index to represent the pest activity level;

[0012] S103: Quantify the environmental stress based on the real-time environmental parameters to obtain an environmental stress coefficient.

[0013] Preferably, the mathematical description of the voiceprint migration index is S m =∫[P(f)·W(f)]df, where P(f) is the sound pressure spectrum and W(f) is the weighting function of pest migration characteristics.

[0014] Preferably, the step S103 specifically includes:

[0015] Acquire the real-time environmental parameters, including real-time temperature and real-time humidity, and obtain the environmental stress coefficient, which is mathematically described as Among them, T is the real-time temperature, T0 is the pest active temperature, H is the real-time humidity, and H0 is the pest active humidity.

[0016] Preferably, the step S2 specifically includes:

[0017] S201: Create a digital farmland map and divide it into n×n grids, and mark each grid with the sensing information;

[0018] S202: Obtaining the pest infection probability through the perception information and constructing the cellular automaton model;

[0019] S203: simulating the pest spread in the next M days based on the cellular automaton model, and outputting the risk heat map based on the digital farmland map.

[0020] Preferably, the calculation formula for the probability of pest infection in step S202 is: Among them, α is the visual impact factor, β is the voiceprint synergy factor, γ is the environmental suppression coefficient, V neighbor is the insect population density, S m is the voiceprint migration index, and E is the environmental stress coefficient.

[0021] Preferably, the step S203-1 specifically includes:

[0022] S203-1: Obtain the current digital farmland map, set a judgment threshold, and mark it as safe grid, risk grid, and high-risk grid according to the insect density;

[0023] S203-2: Predict the probability of the grids adjacent to the risk grid and the high-risk grid being infected by pests M days later based on the cellular automaton model, update the grid labels based on the probability results, and obtain the risk heat map.

[0024] Preferably, the step S3 specifically includes:

[0025] S301: Construct a set of optional measures and mark the effect index, cost index and effective days;

[0026] S302: Inputting restriction conditions according to the risk heat map and generating feasible combinations;

[0027] S303: Constructing a dual objective function;

[0028] S304: Outputting the pest physical and chemical control gradient decision based on the constraint condition and the dual objective function.

[0029] Preferably, the mathematical description of the dual objective function is Among them, f1 is the effect objective function, f2 is the cost objective function, a i is the measure intensity of feasible combination i, R i is the effect index of feasible combination i, Q i is the cost index of feasible combination i.

[0030] A multimodal perception-driven gradient decision-making system for pest physical and chemical control, used to implement the above-mentioned gradient decision-making method for pest physical and chemical control, including a perception module, a diffusion simulation module, and a gradient decision output module;

[0031] The perception module is used to obtain multimodal perception data and build a multimodal perception model, and output perception information, which includes insect population density, soundprint migration index and environmental stress coefficient;

[0032] The diffusion simulation module is used to construct a cellular automation model based on the perception information, simulate the spread of pests and output a risk heat map;

[0033] The gradient decision output module is used to output a gradient decision on the physical and chemical control of pests based on the risk heat map.

[0034] The beneficial effects of the present invention are:

[0035] (1) Improving perception accuracy through three-dimensional quantification of pest behavior characteristics: reducing the error of the soundprint migration index through the weighted function of sound pressure spectrum and migration characteristics; based on the nonlinear environmental stress model, the pest outbreak threshold under extreme climate conditions can be accurately predicted.

[0036] (2) Improve the accuracy of pest prediction through cellular automaton optimization: introduce the infection probability formula to improve the accuracy of pest spread path prediction; generate a risk heat map to intuitively display the risk level.

[0037] (3) Improve control efficiency through gradient control optimization: The dual objective function improves control effectiveness while reducing costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] To facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0039] Figure 1 This is a flowchart of the steps of the gradient decision-making method for pest physical and chemical control based on multimodal perception drive of the present invention. DETAILED DESCRIPTION

[0040] In order to better understand the present invention, various aspects of the present invention will be described in more detail with reference to the accompanying drawings. It should be understood that these detailed descriptions are only descriptions of exemplary embodiments of the present invention and are not intended to limit the scope of the present invention in any way. Throughout the specification, the expression "and / or" includes any and all combinations of one or more of the associated listed items. As used herein, the terms "roughly", "approximately" and similar terms are used as terms to indicate approximate values, rather than as terms to indicate degree, and are intended to illustrate inherent deviations in measurements or calculated values ​​that will be recognized by those of ordinary skill in the art. In addition, in the present invention, the order in which the steps are described does not necessarily represent the order in which these processes occur in actual operation, unless otherwise specified or can be derived from the context.

[0041] It should also be understood that expressions such as "comprises," "including," "having," "includes," and / or "comprising" are open rather than closed expressions in this specification, indicating the presence of the stated features, elements, and / or components, but do not exclude the presence of one or more other features, elements, components, and / or combinations thereof. In addition, when expressions such as "at least one of..." appear after a list of listed features, they modify the entire list of features rather than just the individual elements in the list. In addition, when describing embodiments of the present invention, "may" is used to mean "one or more embodiments of the present invention." And, the term "exemplary" is intended to refer to an example or illustration.

[0042] Unless otherwise defined, all terms used herein (including engineering terms and scientific and technological terms) have the same meaning as commonly understood by those skilled in the art to which the present invention pertains. It should also be understood that, unless otherwise expressly stated in the present invention, words defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an idealized or overly formal sense.

[0043] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments of the present invention can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0044] Example 1:

[0045] See also Figure 1 , a gradient decision-making method for pest physical and chemical control based on multimodal perception, including:

[0046] S1: Acquire multimodal perception data and build a multimodal perception model to output perception information, including insect population density, soundprint migration index, and environmental stress coefficient;

[0047] S2: constructing a cellular automation model based on the perception information, simulating the spread of pests and outputting a risk heat map;

[0048] S3: Outputting a gradient decision on the physical and chemical pest control based on the risk heat map.

[0049] In this embodiment, multimodal perception data is acquired, a multimodal perception model is constructed, and perception information is output. Specifically, the following steps can be performed:

[0050] S101: Uses visual sensors, combined with AI software and machine learning methods to classify and count pests, and outputs insect density (e.g., 50 / m 2 ) and the rate of change of insect abundance (e.g. +200% means the insect abundance has tripled); Example: Yesterday, 20 insects / m were found in field 1. 2 Today it becomes 60 / m 2 , then the insect population change rate = (60-20) / 20 = 200%, and the insect population triples.

[0051] S102: Capture the pest activity sound through the voiceprint sensor, and extract the characteristic frequency through Mel spectrum convolution (for example, the locust flying sound is 80-120Hz), and define the voiceprint migration index to represent the pest activity level. The larger the voiceprint migration index value, the more active the pest. The mathematical description of the voiceprint migration index is S m=∫[P(f)·W(f)]df, where P(f) is the sound pressure spectrum and W(f) is the weighting function of pest migration characteristics. For example, when locusts migrate, W(80-120 Hz) = 3.0;

[0052] S103: Obtain real-time temperature and real-time humidity, obtain environmental stress coefficient, and quantify environmental stress. The mathematical description of the environmental stress coefficient is: Where T is the current temperature, T0 is the pest activity temperature (typically 25°C), H is the current humidity, and H0 is the pest activity humidity (typically 70%). A positive environmental stress coefficient indicates a suitable environment for pest activity, while a negative value indicates a suppressed environment. For example, if the current temperature is 30°C and the humidity is 80%, then E = e^(5 / 10) × ln(80 / 70) = 0.22, indicating suitable conditions for pest activity.

[0053] In this embodiment, a cellular automaton model is constructed based on the perception information to simulate pest spread and output a risk heat map, which can be specifically implemented by the following steps:

[0054] S201: Creating a digital farmland map and dividing it into n×n grids, annotating each grid with the sensing information, that is, annotating each grid with insect population density, soundprint migration index, and environmental stress coefficient;

[0055] S202: Obtain the pest infection probability through the perception information and construct the cellular automaton model. The basic rule is: if a grid has pests, the infection probability of the adjacent grids depends on: the number of pests in the current grid (the more the easier it is to spread); whether the environment is suitable for survival; whether the pests are in a migratory state (determined by the voiceprint migration index). The calculation formula for the pest infection probability is: Among them, α is the visual impact factor, with an initial value of 0.3, β is the voiceprint synergy factor, with an initial value of 0.2, and γ is the environmental inhibition coefficient, with an initial value of 0.15; Example: Insect density of a grid = 100, voiceprint migration index = 200, environmental stress coefficient = 0.2, then the probability of insect infection P = 1 / (1+e -(0.3×100+0.2×200-0.15×0.2) )≈100%, indicating that the adjacent grids are definitely infected;

[0056] S203: simulating the pest spread in the next M days based on the cellular automaton model, and outputting the risk heat map based on the digital farmland map.

[0057] S203-1: Obtain the current digital farmland map, set a judgment threshold, and mark it as safe grid, risk grid, and high-risk grid according to the insect density;

[0058] S203-2: Predict the probability that the eight adjacent grids of the risk grid and the high-risk grid will be infected by pests M days later based on the cellular automaton model, and update the grid markings based on the probability results, that is, remark all grids in the current digital farmland map to obtain the risk heat map.

[0059] In this embodiment, outputting a pest physical and chemical control gradient decision based on the risk heat map can be implemented through the following steps:

[0060] S301: Construct a set of optional measures and label the effect index, cost index, and effective days. For example: Measure A: Pheromone interference, effect index 4 (full score 10), cost index 50 yuan / mu, effective days 3; Measure B: Biological control, effect index 7, cost index 120 yuan / mu, effective days 5; Measure C: Physical isolation, effect index 6, cost index 80 yuan / mu, effective days 2; Measure D: Targeted spraying, effect index 9, cost index 200 yuan / mu, effective days 1;

[0061] S302: Input constraints based on the risk heat map and generate feasible combinations. For example, for a high-risk grid, input constraints: pest risk must be reduced by at least 60%; total cost does not exceed 300 yuan / mu; for a risk grid, input constraints: pest risk must be reduced by at least 40%; total cost does not exceed 100 yuan / mu; generate all feasible combinations: Combination A: targeted pesticide application for 30% of the area + biological control for 20% of the area; Combination B: pheromone for 50% of the area + physical isolation for 40% of the area, etc.

[0062] S303: Constructing a dual objective function Among them, f1 is the effect objective function, f2 is the cost objective function, a i is the measure intensity of feasible combination i, R i is the effect index of feasible combination i, Qi is the cost index of feasible combination i;

[0063] S304: Outputting the pest physical and chemical control gradient decision based on the constraint conditions and the dual objective function, that is, selecting the optimal feasible combination to perform pest control on grids with different risks.

[0064] Example 2:

[0065] A gradient decision-making system for pest physical and chemical control based on multimodal perception drive includes a perception module, a diffusion simulation module and a gradient decision output module.

[0066] The perception module is used to obtain multimodal perception data and build a multimodal perception model, and output perception information, which includes insect density, soundprint migration index and environmental stress coefficient. Specifically, the perception module uses visual sensors, combined with AI software and machine learning methods to classify and count pests, and output insect density and insect population change rate; uses soundprint sensors to capture pest activity sounds, and extracts characteristic frequencies through Mel spectrum convolution, defining a soundprint migration index to characterize the degree of pest activity. A larger value of the soundprint migration index indicates more active pests; and obtains real-time temperature and real-time humidity to obtain an environmental stress coefficient to quantify environmental stress.

[0067] The diffusion simulation module is used to construct a cellular automaton model based on the perception information, simulate the spread of insect pests, and output a risk heat map. Specifically, the module comprises the following steps: establishing a digital farmland map and dividing it into n×n grids, annotating the perception information for each grid, i.e., annotating the insect population density, voiceprint migration index, and environmental stress coefficient for each grid; obtaining the probability of insect pest infection through the perception information and constructing the cellular automaton model. The basic rule is: if an insect pest occurs in one grid, the infection probability of the adjacent grids depends on: the current insect population in the grid; whether the environment is suitable for survival; and whether the pests are in a migratory state; simulating the insect pest spread in the next M days based on the cellular automaton model, and outputting the risk heat map based on the digital farmland map; obtaining the current digital farmland map, setting a judgment threshold, and marking the grids as safe grids, risk grids, and high-risk grids according to the insect population density; predicting the probability of the eight grids adjacent to the risk grid and the high-risk grid being infected by insect pests after M days based on the cellular automaton model, and updating the grid labels based on the probability results, i.e., relabeling all grids in the current digital farmland map to obtain the risk heat map.

[0068] The gradient decision output module is used to output the gradient decision of pest physical and chemical control based on the risk heat map, specifically: constructing a set of optional measures and marking the effect index, cost index and effective days; inputting the restriction conditions according to the risk heat map and generating feasible combinations; constructing a dual-objective function; outputting the gradient decision of pest physical and chemical control based on the restriction conditions and the dual-objective function, that is, selecting the optimal feasible combination to carry out pest control on grids with different risks.

[0069] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as a preferred embodiment as above, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications to equivalent embodiments using the technical contents disclosed above without departing from the scope of the technical solution of the present invention. However, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A multimodal perception-driven gradient decision-making method for pest control, characterized by: The implementation of the gradient decision-making method for the physical and chemical pest control includes the following steps: S1: Acquire multimodal perception data and build a multimodal perception model to output perception information, including insect population density, soundprint migration index, and environmental stress coefficient; S2: constructing a cellular automation model based on the perception information, simulating the spread of pests and outputting a risk heat map; S3: Outputting a gradient decision on the physical and chemical pest control based on the risk heat map.

2. The gradient decision-making method for pest physical and chemical control according to claim 1, characterized in that: The step S1 specifically includes: S101: Outputting insect population density and insect population change rate through visual sensors; S102: Capture the sounds of pest activity, extract characteristic frequencies, and define a soundprint migration index to represent the pest activity level; S103: Quantify the environmental stress based on the real-time environmental parameters to obtain an environmental stress coefficient.

3. The gradient decision-making method for pest physical and chemical control according to claim 2, characterized in that: The mathematical description of the voiceprint migration index is S m =∫[P(f)·W(f)]df, where P(f) is the sound pressure spectrum and W(f) is the weighting function of pest migration characteristics.

4. The gradient decision-making method for pest physical and chemical control according to claim 2, characterized in that: The step S103 specifically includes: The real-time environmental parameters are obtained, including real-time temperature and real-time humidity, and the environmental stress coefficient is obtained. The mathematical description of the environmental stress coefficient is: Among them, T is the real-time temperature, T0 is the pest active temperature, H is the real-time humidity, and H0 is the pest active humidity.

5. The gradient decision-making method for pest physical and chemical control according to claim 1, characterized in that: The step S2 specifically includes: S201: Create a digital farmland map and divide it into n×n grids, and mark each grid with the sensing information; S202: Obtaining the pest infection probability through the perception information and constructing the cellular automaton model; S203: simulating the pest spread in the next M days based on the cellular automaton model, and outputting the risk heat map based on the digital farmland map.

6. The gradient decision-making method for pest physical and chemical control according to claim 5, characterized in that: The calculation formula for the pest infection probability in step S202 is: Among them, α is the visual impact factor, β is the voiceprint synergy factor, γ is the environmental suppression coefficient, V neighbor is the insect population density, S m is the voiceprint migration index, and E is the environmental stress coefficient.

7. The gradient decision-making method for pest physical and chemical control according to claim 5, characterized in that: The step S203-1 specifically includes: S203-1: Obtain the current digital farmland map, set a judgment threshold, and mark it as safe grid, risk grid, and high-risk grid according to the insect density; S203-2: Predict the probability of the grids adjacent to the risk grid and the high-risk grid being infected by pests M days later based on the cellular automaton model, update the grid labels based on the probability results, and obtain the risk heat map.

8. The gradient decision-making method for pest physical and chemical control according to claim 1, characterized in that: The step S3 specifically includes: S301: Construct a set of optional measures and mark the effect index, cost index and effective days; S302: Inputting restriction conditions according to the risk heat map and generating feasible combinations; S303: Constructing a dual objective function; S304: Outputting the pest physical and chemical control gradient decision based on the constraint condition and the dual objective function.

9. The gradient decision-making method for pest physical and chemical control according to claim 8, characterized in that: The mathematical description of the dual objective function is Among them, f1 is the effect objective function, f2 is the cost objective function, a i is the measure intensity of feasible combination i, R i is the effect index of feasible combination i, Q i is the cost index of feasible combination i.

10. A multimodal perception-driven pest control gradient decision-making system, characterized by: The system is applied to the gradient decision-making method for physical and chemical pest control as described in any one of claims 1 to 9, comprising a perception module, a diffusion simulation module and a gradient decision output module; The perception module is used to obtain multimodal perception data and build a multimodal perception model, and output perception information, which includes insect population density, soundprint migration index and environmental stress coefficient; The diffusion simulation module is used to construct a cellular automation model based on the perception information, simulate the spread of pests and output a risk heat map; The gradient decision output module is used to output a gradient decision on the physical and chemical control of pests based on the risk heat map.

Images