Prediction method of pest population density distribution pattern in deciduous forest

Abstract

The invention relates to the technical field of deciduous forest pest prediction, and discloses a method for predicting pest population density distribution patterns in deciduous forests. The method comprises the following steps: step 1, collecting historical information of pest density at a predicted local point; step 2, predicting a spatial distribution pattern according to the historical information of pest density in the local area; step 3, when data are distributed in two different intervals, xl and xh are respectively used to represent the midpoints of the two intervals, and s=xh-xl represents the distance between the two midpoints; step 4, Ul and Uh are respectively used to represent half of the width of the two intervals; step 5, the midpoints xl and xh of the two intervals are regarded as two states, and Ul and Uh respectively represent the uncertainty of the two states; step 6, calculating K=(Ul+Uh) / s, and the spatial distribution structure of the pests can be obtained according to the value of K.

Description

Technical Field

[0001] The invention relates to the technical field of deciduous forest pest prediction, in particular to a method for predicting pest population density distribution patterns in deciduous forests. Background Art

[0002] Insects are typical cold-blooded animals. Temperature is a key ecological factor that affects insect growth and development and population dynamics. Climate warming will affect insect growth and development, geographical distribution, and population density in different temperature zones. Accurate pest prediction and forecasting is a prerequisite for scientific pest control.

[0003] The application of prediction models in pest prediction is closely related to the development of mathematical methods. With the development of mathematical theory and other disciplines, pest prediction has roughly gone through four stages of development: empirical prediction, experimental prediction, statistical prediction and information prediction.

[0004] At present, the existing technology for pest prediction methods is a pest prediction method and system based on a population model, such as the Chinese patent publication number (CN110110945B), which inputs the initial number of pupae and the average daily temperature of all days into the spring emergence model to obtain the number of adults emerging on the first day; inputs the number of adults in the previous n-1 days and the average daily temperature of the nth day into the adult egg-laying model to obtain the number of eggs laid by adults on the nth day; inputs the number of eggs laid in the previous n-1 days and the average daily temperature of the nth day into the egg development model to obtain the number of larvae developed on the nth day; inputs the number of larvae in the previous n-1 days and the average daily temperature of the nth day into the larval development model to obtain the number of pupae developed on the nth day; inputs the number of pupae in the previous n-1 days and the average daily temperature of the nth day into the pupal development model to obtain the number of adults emerging on the nth day; judges whether n is equal to the total number of days; if so, outputs the corresponding number of adults, number of eggs laid, number of larvae and number of pupae for each day. The present invention is not limited to a specific area and can achieve long-term prediction of pests in multiple areas.

[0005] The above-mentioned technical solution is to select local insect pupae and local temperature through field surveys for modeling, then simulate the pest breeding process, and finally obtain the density data of the local pest population. However, this solution is only suitable for alien invasive species that have no natural enemies. The local pest population is restricted by the food chain and the local vegetation, water and soil environment, and it is difficult to have a uniform exponential rising data distribution. Summary of the invention

[0006] 1. Technical issues to be resolved

[0007] In view of the shortcomings of the existing technology, the present invention provides a method for predicting the density distribution pattern of pest populations in deciduous forests. It has the advantages of not requiring field exploration to capture pupae for breeding experiments to reduce prediction costs, and can predict the density distribution data of native pests in future years, thereby solving the problems of the above-mentioned technologies.

[0008] (II) Technical solution

[0009] To achieve the above object, the present invention provides the following technical solution: a method for predicting the distribution pattern of pest population density in a deciduous forest, comprising the following steps:

[0010] Step 1: Collect and predict the pest density of local points over the years;

[0011] Step 2: predicting the spatial distribution pattern based on the historical information of pest density in the local area;

[0012] Step 3: When the data is distributed in two different intervals, let xl and xh represent the midpoints of the two intervals respectively, then s=xh-xl represents the distance between the two midpoints;

[0013] Step 4: Let Ul and Uh represent half of the width of the two intervals respectively;

[0014] Step 5: Consider the midpoints xl and xh of the two intervals as two states, and Ul and Uh represent the uncertainty of the two states respectively;

[0015] Step 6. Calculate K = (Ul + Uh) / s. The spatial distribution structure of pests can be obtained based on the value of K.

[0016] Preferably, the historical information range of pest density at the local point includes the pest density information table of the local point in the past ten years, or the pest density information table of the local point that is recorded and actually exists for one year or more than one year.

[0017] Through the above technical scheme, by collecting the pest density information table of the local point in the area to be predicted, the closer the pest density information table is to the current year and the more consistent the relationship between the pest density information tables of previous years, the higher the completeness of the information. Otherwise, the recorded pest density information table is collected for modeling calculation. Since the prediction operation is directly based on the pest density information table, and the pest density information table of the area is the most realistic feedback of the pest density information of the area in previous years, the need to actually capture pupae and cultivate them in the area is avoided, which causes economic expenditure. The advantage of using the pest density information table is that the localized pests have adapted to the local ecological environment. The pests existing in the local area will only be insects that adapt to the local environment. There will be no exponential reproduction caused by the lack of natural enemy inhibition to form variables that disturb the prediction results.

[0018] Preferably, the distribution pattern prediction includes normal distribution, Poisson distribution and exponential distribution.

[0019] Through the above technical solution, multiple distribution mode predictions are made to ensure the accuracy of the prediction results.

[0020] Preferably, the distribution pattern prediction further comprises the following steps:

[0021] S1. Collect and predict the historical information data of local pest density to ensure the accuracy and completeness of the data;

[0022] S2. Conduct exploratory analysis on the collected pest density data, and understand the distribution characteristics, outliers and trends of the data by drawing histograms and box plot visualization methods;

[0023] S3. Based on the analysis of the data, select a suitable probability distribution model to describe the distribution characteristics of the data;

[0024] S4, performing parameter estimation on the selected distribution model, and determining the parameter values ​​in the model using maximum likelihood estimation and Bayesian method;

[0025] S5. Evaluate the fit and predictive ability of the selected distribution model by comparing it with the actual data;

[0026] S6. Use the fitted model to predict the distribution pattern of future events or phenomena based on future forecast data.

[0027] Preferably, the t represents the time unit year.

[0028] Preferably, the x represents the local density of pests.

[0029] Preferably, when the calculation result K of K=(Ul+Uh) / s is less than 0.75, the spatial distribution of pests is a target-like structure diagram.

[0030] Through the above technical solution, the target structure diagram is orderly distributed, so it can be predicted that the distribution of pests in the area next year will be a normal uniform distribution.

[0031] Preferably, when the calculation result K of K=(Ul+Uh) / s is in the interval [0.75, 0.85], the spatial distribution of pests is an irregular patch structure diagram.

[0032] Through the above technical solution, it can be directly understood from the irregular patch structure diagram that the pest density distribution in the area next year is relatively concentrated, showing an irregular concentrated distribution trend.

[0033] Preferably, when the calculation result K of K=(Ul+Uh) / s is in the interval (0.85,0.91], the edge of the spatial distribution structure diagram of pests becomes blurred.

[0034] Through the above technical solution, it can be directly understood from the edge fuzzy structure diagram that the pest density distribution in the area next year is a chaotic distribution trend and the distribution is disordered.

[0035] Preferably, when the calculation result of K=(Ul+Uh) / s is greater than 0.91, the spatial distribution pattern is chaotic.

[0036] Through the above technical solution, it can be directly understood through the chaotic distribution pattern that the density distribution of pests in the area next year is completely disordered and chaotic and the density is extremely high, which shows that the pests in the area next year will cause great harm.

[0037] Compared with the prior art, the present invention provides a method for predicting the distribution pattern of pest population density in deciduous forests, which has the following beneficial effects:

[0038] 1. The prediction method of the pest population density distribution pattern in the deciduous forest uses the advantage of the pest density information table that the local pests have adapted to the local ecological environment. The local pests will only be the insects that adapt to the local environment. The pest density information table of the local point to be predicted is collected. The closer the pest density information table is to this year and the more consistent the relationship between the pest density information tables of previous years, the higher the completeness of the information. Otherwise, the recorded pest density information table is collected for modeling calculation. Since the prediction operation is directly based on the pest density information table, and the pest density information table of the area is the most realistic feedback of the pest density information of the area in previous years, the beneficial effect of reducing the prediction cost is achieved without the need for field exploration to capture pupae for breeding experiments.

[0039] 2. The prediction method of the pest population density distribution pattern in the deciduous forest is based on the calculation result of K=(Ul+Uh) / s. When K is less than 0.75, the spatial distribution of pests is a target structure. When K is in the interval [0.75, 0.85], the spatial distribution of pests is an irregular patch structure. When K is in the interval (0.85, 0.91], the edge of the spatial distribution structure of pests becomes blurred. When K is greater than 0.91, the spatial distribution pattern is chaotic, which achieves the beneficial effect of predicting the density distribution data of native pests in the future year. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a schematic diagram of the process of the present invention;

[0041] Figure 2 This is a schematic diagram of the data model when the K calculation result of the present invention is less than 0.75;

[0042] Figure 3 This is a schematic diagram of the data model when the K operation result of the present invention is in the interval [0.75, 0.85];

[0043] Figure 4 Schematic diagram of the data model when the K operation result of the present invention is in the interval (0.85, 0.91];

[0044] Figure 5This is a schematic diagram of the data model when the K calculation result of the present invention is greater than 0.91. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] See also Figure 1-5 A method for predicting the distribution pattern of pest population density in a deciduous forest comprises the following steps:

[0047] Step 1: Collect and predict the pest density of local points over the years;

[0048] Step 2: predict the spatial distribution pattern based on the historical information of pest density in the local area;

[0049] Step 3: When the data is distributed in two different intervals, let xl and xh represent the midpoints of the two intervals respectively, then s=xh-xl represents the distance between the two midpoints;

[0050] Step 4: Let Ul and Uh represent half of the width of the two intervals respectively;

[0051] Step 5: Consider the midpoints xl and xh of the two intervals as two states, and Ul and Uh represent the uncertainty of the two states respectively;

[0052] Step 6. Calculate K = (Ul + Uh) / s. The spatial distribution structure of pests can be obtained based on the value of K.

[0053] Specifically, the historical information on pest density at a local point includes the pest density information table for the local point in the past ten years, or the pest density information table for the local point that is recorded and actually exists for one year or more.

[0054] The advantage is that by collecting the local point pest density information table that needs to be predicted, the closer the pest density information table is to this year and the more consistent the relationship between the pest density information tables of previous years, the higher the completeness of the information. Otherwise, the recorded pest density information table is collected for modeling calculation. Since the prediction operation is directly based on the pest density information table, and the pest density information table of the area is the most realistic feedback of the pest density information of the area in previous years, the need to actually capture pupae and cultivate them in the area is avoided, which causes economic expenditure. The advantage of using the pest density information table is that the localized pests have adapted to the local ecological environment. The pests that exist locally will only be insects that adapt to the local environment. There will be no exponential reproduction caused by the lack of natural enemy inhibition to form variables that disturb the prediction results.

[0055] Specifically, distribution pattern predictions include normal distribution, Poisson distribution, and exponential distribution.

[0056] The advantage is that the accuracy of the prediction results is ensured by predicting through multiple distribution patterns.

[0057] Specifically, the distribution pattern prediction also includes the following steps:

[0058] S1. Collect and predict the historical information data of local pest density to ensure the accuracy and completeness of the data;

[0059] S2. Conduct exploratory analysis on the collected pest density data, and understand the distribution characteristics, outliers and trends of the data by drawing histograms and box plot visualization methods;

[0060] S3. Based on the analysis of the data, select a suitable probability distribution model to describe the distribution characteristics of the data;

[0061] S4, performing parameter estimation on the selected distribution model, and determining the parameter values ​​in the model using maximum likelihood estimation and Bayesian method;

[0062] S5. Evaluate the fit and predictive ability of the selected distribution model by comparing it with the actual data;

[0063] S6. Use the fitted model to predict the distribution pattern of future events or phenomena based on future forecast data.

[0064] Specifically, t represents the time unit year.

[0065] Specifically, x represents the local density of pests.

[0066] Please refer to the instruction manual Figure 2Specifically, when K=(Ul+Uh) / s and the calculation result K is less than 0.75, the spatial distribution of pests is a target-shaped structure diagram. The advantage is that through the orderly distribution of the target-shaped structure diagram, it can be predicted that the distribution of pests in the area next year will be a normal uniform distribution.

[0067] Please refer to the instruction manual Figure 3 Specifically, when K=(Ul+Uh) / s and the calculation result K is in the interval [0.75,0.85], the spatial distribution of pests is an irregular patch structure diagram. The advantage is that through the irregular patch structure diagram, it can be directly understood that the density distribution of pests in the area next year is relatively concentrated, showing an irregular concentrated distribution trend.

[0068] Please refer to the instruction manual Figure 4 Specifically, when K=(Ul+Uh) / s is in the interval (0.85,0.91], the edge of the spatial distribution structure diagram of pests becomes blurred. The advantage is that through the edge fuzzy structure diagram, it can be directly understood that the density distribution of pests in the area next year is a chaotic distribution trend and the distribution is disordered.

[0069] Please refer to the instruction manual Figure 5 Specifically, when the calculated result of K=(Ul+Uh) / s is greater than 0.91, the spatial distribution pattern is chaotic. The advantage is that the chaotic distribution pattern can directly show that the density distribution of pests in the area next year is completely disordered and chaotic with extremely high density, indicating that the pests in the area will cause greater damage next year.

[0070] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for predicting the distribution pattern of pest population density in a deciduous forest, characterized in that: The following steps are involved: Step 1: Collect and predict the pest density of local points over the years; Step 2: predicting the spatial distribution pattern based on the historical information of pest density in the local area; Step 3: When the data is distributed in two different intervals, let xl and xh represent the midpoints of the two intervals respectively, then s=xh-xl represents the distance between the two midpoints; Step 4: Let Ul and Uh represent half of the width of the two intervals respectively; Step 5: Consider the midpoints xl and xh of the two intervals as two states, and Ul and Uh represent the uncertainty of the two states respectively; Step 6: Calculate K = (Ul + Uh) / s. The spatial distribution structure of pests can be obtained based on the value of K. The historical information of pest density at the local point includes the pest density information table of the local point in the last ten years, or the pest density information table of the local point that is recorded and actually exists for one year or more; The distribution pattern prediction includes normal distribution, Poisson distribution and exponential distribution; The distribution pattern prediction further comprises the following steps: S1. Collect and predict the historical information data of local pest density to ensure the accuracy and completeness of the data; S2. Conduct exploratory analysis on the collected pest density data, and understand the distribution characteristics, outliers and trends of the data by drawing histograms and box plot visualization methods; S3. Based on the analysis of the data, select a suitable probability distribution model to describe the distribution characteristics of the data; S4, performing parameter estimation on the selected distribution model, and determining the parameter values ​​in the model using maximum likelihood estimation and Bayesian method; S5. Evaluate the fit and predictive ability of the selected distribution model by comparing it with the actual data; S6. Use the fitted model to predict the distribution pattern of future events or phenomena based on future forecast data; Said t represents the time unit year; The x represents the local density of pests; When the calculation result K=(Ul+Uh) / s is less than 0.75, the spatial distribution of pests is a target-shaped structure diagram; When the calculation result K=(Ul+Uh) / s is in the interval [0.75,0.85], the spatial distribution of pests is an irregular patch structure diagram; When the calculation result K=(Ul+Uh) / s is in the interval (0.85,0.91], the edge of the spatial distribution structure diagram of pests becomes blurred; When the calculated result of K=(Ul+Uh) / s is greater than 0.91, the spatial distribution pattern is chaotic.

Images