CN111445062A

CN111445062A - Insect pest position prediction method and device, computer equipment and storage medium

Info

Publication number: CN111445062A
Application number: CN202010184911.XA
Authority: CN
Inventors: 尹海波; 马啸
Original assignee: Shenzhen Shuliantianxia Intelligent Technology Co Ltd
Current assignee: Shenzhen Shuliantianxia Intelligent Technology Co Ltd
Priority date: 2020-03-17
Filing date: 2020-03-17
Publication date: 2020-07-24
Anticipated expiration: 2040-03-17
Also published as: CN111445062B

Abstract

The embodiment of the invention discloses a method and a device for predicting a pest position, computer equipment and a storage medium, wherein the method comprises the following steps: acquiring a historical pest set, wherein the historical pest set comprises a plurality of pest positions and a first pest time corresponding to each pest position; obtaining pest reference positions according to the multiple pest positions; obtaining a category central position corresponding to a target azimuth category at each first pest time according to each pest position, the first pest time corresponding to each pest position and the pest reference position; and obtaining a target pest position prediction result according to the pest reference position and the category central position corresponding to the target azimuth category at each first pest time. By the method, the pest position can be effectively predicted.

Description

Insect pest position prediction method and device, computer equipment and storage medium

Technical Field

The present invention relates to the field of pest location prediction technologies, and in particular, to a method and an apparatus for predicting a pest location, a computer device, and a storage medium.

Background

The monitoring of crop diseases and insect pests is the basis for controlling the damage of diseases and insects and ensuring the high quality and high yield of crops. With the continuous development of computer technology in various industries, the computer technology is also widely applied to the field of plant protection, and a large number of databases and management systems related to plant diseases and insect pests have been developed at home and abroad.

At present, most of the systems aim at pest monitoring, and the types and the severity of pests are identified by technologies such as computer vision and the like. For example, the positions of the pests are obtained through a GPS (global positioning system) arranged in the mobile phone, and the types and the severity of the pests are confirmed through pest images shot by a camera of the mobile phone.

Although these systems can monitor the pest data, how to predict the location of the pest, so as to effectively kill the pest on the cradle, is of great significance to agricultural production.

Disclosure of Invention

In view of the above, it is necessary to provide a pest location prediction method, device, computer device, and storage medium for effectively predicting a location of a pest.

In a first aspect, a method for predicting pest location is provided, the method comprising: acquiring a historical pest set, wherein the historical pest set comprises a plurality of pest positions and a first pest time corresponding to each pest position; obtaining pest reference positions according to the multiple pest positions; obtaining a category central position corresponding to a target azimuth category at each first pest time according to each pest position, a first pest time corresponding to each pest position and the pest reference position, wherein the category central position is the center of the pest position within a target azimuth range, the target azimuth range is an azimuth range corresponding to the target azimuth category, the target azimuth range is any one azimuth range of a plurality of preset azimuth ranges, and the plurality of azimuth ranges are overlapped to cover all azimuths; and obtaining a target pest position prediction result according to the pest reference position and the category central position corresponding to the target azimuth category at each first pest time, wherein the target pest position prediction result is a prediction result of a pest position within a target azimuth angle range at a second pest time when the pest reference position is used as the reference position, and the second pest time is a time after the first pest time.

In a second aspect, a pest location prediction device is provided, comprising: the pest control system comprises a position acquisition module, a pest control module and a pest control module, wherein the position acquisition module is used for acquiring a historical pest set which comprises a plurality of pest positions and first pest time corresponding to each pest position; the reference determining module is used for obtaining pest reference positions according to the multiple pest positions; the category determination module is used for obtaining a category central position corresponding to a target azimuth category at each first pest attack time according to each pest attack position, a first pest attack time corresponding to each pest attack position and the pest attack reference position, wherein the category central position is the center of the pest attack position within a target azimuth angle range, the target azimuth angle range is an azimuth angle range corresponding to the target azimuth category, the target azimuth angle range is any one azimuth angle range of a plurality of preset azimuth angle ranges, and the plurality of azimuth angle ranges are overlapped to cover all azimuths; and the result prediction module is used for obtaining a target pest position prediction result according to the pest reference position and the category central position corresponding to the target azimuth category at each first pest time, wherein the target pest position prediction result is a prediction result of a pest position within a target azimuth angle range at a second pest time when the pest reference position is used as the reference position, and the second pest time is a time after the first pest time.

In a third aspect, there is provided a computer device comprising a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of: acquiring a historical pest set, wherein the historical pest set comprises a plurality of pest positions and a first pest time corresponding to each pest position; obtaining pest reference positions according to the multiple pest positions; obtaining a category central position corresponding to a target azimuth category at each first pest time according to each pest position, a first pest time corresponding to each pest position and the pest reference position, wherein the category central position is the center of the pest position within a target azimuth range, the target azimuth range is an azimuth range corresponding to the target azimuth category, the target azimuth range is any one azimuth range of a plurality of preset azimuth ranges, and the plurality of azimuth ranges are overlapped to cover all azimuths; and obtaining a target pest position prediction result according to the pest reference position and the category central position corresponding to the target azimuth category at each first pest time, wherein the target pest position prediction result is a prediction result of a pest position within a target azimuth angle range at a second pest time when the pest reference position is used as the reference position, and the second pest time is a time after the first pest time.

In a fourth aspect, there is provided a computer readable storage medium storing a computer program which, when executed by a processor, causes the processor to perform the steps of: acquiring a historical pest set, wherein the historical pest set comprises a plurality of pest positions and a first pest time corresponding to each pest position; obtaining pest reference positions according to the multiple pest positions; obtaining a category central position corresponding to a target azimuth category at each first pest time according to each pest position, a first pest time corresponding to each pest position and the pest reference position, wherein the category central position is the center of the pest position within a target azimuth range, the target azimuth range is an azimuth range corresponding to the target azimuth category, the target azimuth range is any one azimuth range of a plurality of preset azimuth ranges, and the plurality of azimuth ranges are overlapped to cover all azimuths; and obtaining a target pest position prediction result according to the pest reference position and the category central position corresponding to the target azimuth category at each first pest time, wherein the target pest position prediction result is a prediction result of a pest position within a target azimuth angle range at a second pest time when the pest reference position is used as the reference position, and the second pest time is a time after the first pest time.

The embodiment of the invention has the following beneficial effects:

the invention provides a method, a device, computer equipment and a storage medium for predicting pest positions. Moreover, the prediction result of the pest position is the prediction result in the target azimuth angle range, that is, the prediction results in different azimuth angle ranges can be obtained in the above manner, and compared with the manner that only one prediction result at the second pest time can be obtained in a prediction manner, the manner can obtain a more precise prediction result.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

Wherein:

FIG. 1 is a schematic diagram of a process for implementing a pest location prediction method in one embodiment;

FIG. 2 is a diagram of azimuth categories and azimuth ranges in one embodiment;

FIG. 3 is a flow diagram illustrating an implementation of step 106 in one embodiment;

FIG. 4 is a schematic illustration of pest reference location and category center location for one embodiment;

FIG. 5 is a schematic diagram of a flowchart of an implementation of a pest location prediction method in one embodiment;

FIG. 6 is a flow diagram illustrating an implementation of step 108 in one embodiment;

FIG. 7 is a schematic illustration of the effect of wind vectors in one embodiment;

FIG. 8 is a block diagram of a pest location prediction device according to one embodiment;

FIG. 9 is a block diagram of a computer device in one embodiment.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

In one embodiment, a pest location prediction method is provided, and a main execution subject of the pest location prediction method according to the embodiment of the present invention is a device capable of implementing the pest location prediction method according to the embodiment of the present invention, and the device may include, but is not limited to, a terminal and a server. The terminal comprises a desktop terminal and a mobile terminal, wherein the desktop terminal comprises but is not limited to a desktop computer and a vehicle-mounted computer; mobile terminals include, but are not limited to, cell phones, tablets, laptops, and smartwatches. The server includes a high performance computer and a cluster of high performance computers.

As shown in fig. 1, the method for predicting the location of an insect pest according to the embodiment of the present invention specifically includes:

102, obtaining a historical pest set, wherein the historical pest set comprises a plurality of pest positions and first pest time corresponding to each pest position.

The pest position is the position of a pest, and can be represented in a form of longitude and latitude; to simplify the representation and facilitate subsequent calculations, pest location may also be represented in the form of π, for example, by equations (1) and (2), where a_iLongitude or latitude corresponding to the position of the ith pest in the historical pest set.

The first pest attack time is time obtained according to a preset rule based on the third pest attack time, wherein the third pest attack time is time for detecting a pest attack position. It is noted that a first pest time corresponds to at least one pest location.

For example, when the time range corresponding to the third pest time is relatively large, for example, when the time range corresponding to the third pest time is one week, the preset rule may be: segmenting the third pest time to obtain a plurality of first pest times, for example, segmenting one third pest time to obtain 7 first pest times; for example, when the time range corresponding to the third pest time is smaller or the pest location corresponding to each third pest time is smaller, for example, the time range corresponding to the third pest time is one day, the preset rule may be: and combining the plurality of third pest attack times to obtain a first pest attack time. The purpose of combining or segmenting the third pest time is mainly to enable the number of pest positions in the finally determined first pest time to be reasonable. How the specific preset rule is, is not specifically limited herein. Table 1 shows pest locations corresponding to a third pest time, and it can be seen that a third pest time may also correspond to multiple pest locations, i.e., multiple pest locations are detected on the same day.

TABLE 1

Two examples of pest locations are provided: firstly, acquiring a pest position through a terminal with a GPS; secondly, set up the pest monitoring point in a plurality of different positions in advance to the longitude and latitude that the pest monitoring point corresponds is recorded in advance, when monitoring the pest at certain pest monitoring point, just with the corresponding longitude and latitude of this pest monitoring point as the pest position.

And 104, acquiring a pest reference position according to the multiple pest positions.

The pest reference position refers to a reference position determined for obtaining a target pest position prediction result, namely, the pest position is predicted according to the relative position relation between the pest reference position and a category center position corresponding to the target azimuth category at each first pest time, and the finally obtained target pest position prediction result takes the pest reference position as a reference.

An example of obtaining a pest reference location is provided, comprising: selecting a part of pest positions from the plurality of pest positions; calculating to obtain a central position corresponding to the partial pest position according to the selected partial pest position; and taking the central position corresponding to the part of pest damage positions as pest damage reference positions.

For example, the selected part of pest positions are: {150 °, 60 ° }, {153 °, 62 ° }, {161 °, 70 ° }, the central position (pest reference position) corresponding to the partial pest position is: { (150 ° +153 ° +161 °)/3, (60 ° +62 ° +70 °)/3 }.

Illustratively, a method of selecting a subset of pest locations from the plurality of pest locations is provided: arranging the first pest damage time according to the time sequence to obtain the first pest damage time with the earliest time; and selecting the pest position corresponding to the first pest time with the earliest time as the partial pest position.

For example, the three first pest times in the historical pest set are 2000.10.01, 2000.10.02 and 2000.10.03, so that the first pest time with the earliest time is 2000.10.01, and each pest position corresponding to 2000.10.01 is selected as a part of pest positions.

And 106, obtaining a category central position corresponding to a target azimuth category at each first pest attack time according to each pest attack position, the first pest attack time corresponding to each pest attack position and the pest attack reference position, wherein the category central position is the center of the pest attack position within a target azimuth angle range, the target azimuth angle range is an azimuth angle range corresponding to the target azimuth category, the target azimuth angle range is any one azimuth angle range in a plurality of preset azimuth angle ranges, and the plurality of azimuth angle ranges are overlapped to cover all azimuths.

The direction type refers to a type of a preset direction for distinguishing insect pests in different directions so as to predict insect pest positions more finely. And the target azimuth category is an azimuth category corresponding to the target azimuth range.

As shown in FIG. 2, all azimuths, meaning azimuths encompassing 0 to 360, include a total of 8 azimuthal ranges in FIG. 2: [0 °, 45 °), [45 °, 90 °), [90 °, 135 °), [135 °, 180 °), [180 °, 225 °), [225 °, 270 °), [270 °, 315 °, and [315 °, 360 ° ], the corresponding orientation categories are category 1, category 2, …, category 8, if the target azimuth range is [270 °, 315 °), the target orientation category is category 7, and the center of the pest position within the target azimuth range is [270 °, 315 °) is a solid small triangle in fig. 2, that is, the solid small triangle is the category center position of category 7.

In one embodiment, an example of determining a category center position is provided, and as shown in fig. 3, the step 106 of obtaining a category center position corresponding to a target azimuth category at each first pest time according to each pest location, a first pest time corresponding to each pest location, and the pest reference location includes:

and 106A, determining an azimuth angle of each pest position relative to the pest reference position according to each pest position and the pest reference position.

The azimuth angle is used for reflecting the relative position relation between the pest position and the pest reference position, and the relative position relation is specifically an angle position relation. Since the latitude and longitude are relative to the spherical earth, the azimuth is calculated using equations (3) to (5).

Wherein, lon_aLongitude value of reference position A of insect pest, lat_aThe latitude value of the insect pest reference position A is obtained; lon_bIs the longitude value of a certain pest location B, lat_bThe latitude value of the pest position B is obtained.

And 106B, determining the pest positions belonging to the target azimuth category according to the azimuth angles of the pest positions relative to the pest reference positions and the target azimuth angle range.

And determining the pest position with the azimuth angle within the target azimuth angle range as the pest position belonging to the target azimuth category. For example, if the azimuth angle of the pest position P1 is 35 °, the azimuth angle of the pest position P2 is 60 °, the target azimuth angle range is [0 °, 45 °), and the target azimuth class is class 1, the pest position belonging to the target azimuth class is P1.

And 106C, obtaining a category central position corresponding to the target azimuth category at each first pest attack time according to the pest attack positions belonging to the target azimuth category based on the first pest attack time corresponding to the pest attack position.

Each first pest time comprises a plurality of azimuth categories, each azimuth category comprises a target azimuth category, in order to obtain a category center position corresponding to the target azimuth category at each first pest time, a pest position belonging to the target azimuth category at the first pest time needs to be found first, then the center of the pest position belonging to the target azimuth category is calculated, and the calculated center is the category center position corresponding to the target azimuth category at the first pest time.

And 108, obtaining a target pest position prediction result according to the pest reference position and the category central position corresponding to the target azimuth category at each first pest time, wherein the target pest position prediction result is a prediction result of a pest position within the target azimuth angle range at a second pest time when the pest reference position is used as the reference position, and the second pest time is a time after the first pest time.

Wherein, second pest time, for the time after a plurality of first pest time in history pest set, for example, 3 first pest time in history pest set are: 2000.10.01, 2000.10.02, and 2000.10.03, the second infestation time can be 2000.10.04.

As shown in fig. 4, from day 1 to day N-1, N-1 first pest damage time is obtained, numerals 1 to 8 represent 8 azimuth categories, solid small dots represent pest damage reference positions, solid small triangles represent category center positions, when a plurality of solid small triangles are in a hollow circle, the solid small triangles represent category center positions corresponding to each azimuth category at a certain first pest damage time, and assuming that day N is second pest damage time, target pest damage position prediction results corresponding to target azimuth categories at day N can be obtained through the steps 102 to 108.

Since the target position category is any one of the plurality of position categories, the pest position prediction results corresponding to other position categories can be obtained through steps 102 to 108.

According to the pest position prediction method, pest reference positions are obtained according to the plurality of pest positions in the historical pest concentration, then according to the plurality of pest positions in the historical pest concentration, the first pest time corresponding to each pest position and the pest reference positions obtain the category center position corresponding to the target position category at each first pest time, and finally, the prediction of the pest positions is achieved according to the pest reference positions and the category center position corresponding to the target position category at each first pest time. Moreover, the prediction result of the pest position is the prediction result in the target azimuth angle range, that is, the prediction results in different azimuth angle ranges can be obtained in the above manner, and compared with the manner that only one prediction result at the second pest time can be obtained in a prediction manner, the manner can obtain a more precise prediction result.

In one embodiment, more refined orientation classification is needed when the number of pests is large, and too refined orientation classification is not needed when the number of pests is small, so that, in order to obtain a classification result with higher accuracy, before obtaining the historical pest set in step 102, as shown in fig. 5, the method further includes:

and step 110, acquiring historical pest quantity in a preset time period.

The preset time period is a preset time period, and can be set according to actual requirements, for example, the preset time period is set to be half a month or even a half year; the historical pest number is the number of pests counted in a preset time period.

And step 112, determining the plurality of azimuth angle ranges according to the historical pest number in the preset time period.

When the number of the historical insect pests is large, the number of azimuth angle ranges is set to be larger, correspondingly, the number of azimuths contained in each azimuth angle range is smaller, so that more precise azimuth classification is carried out, and more accurate prediction results are obtained; when the historical pest number is smaller, the number of the azimuth angle ranges is less, correspondingly, the number of the azimuths contained in each azimuth angle range is larger, and more accurate prediction results can be obtained by containing more pest positions through more azimuths.

Illustratively, a range of standard quantities is obtained [ P1, P2 ]]The standard number of the azimuth angle ranges is M, and when the historical pest number in the preset time period is larger than a large value P2 in the standard number range, the number of the azimuth angle ranges is larger than M; when the historical pest number in the preset time period is smaller than a small value P1 in the standard number range, the number of the azimuth angle ranges is smaller than M; when the historical pest number in the preset time period is in the standard number range [ P1, P2 ]]For example, the number of azimuth angle ranges is m, and a plurality of azimuth angle ranges is determined as (360 degrees × (1-i))/m ≦ depth_iNot more than 360 degrees × i/m, wherein, i is not less than 1 and not more than m, wherein, degree_iIs the ith azimuth range.

Step 114, determining the azimuth category corresponding to each azimuth range.

For example, the azimuth class corresponding to the ith azimuth range is determined as the ith class, but the azimuth class corresponding to the ith azimuth range may also be determined as the mth class, where i ≠ m.

And step 116, determining a target azimuth type corresponding to the target azimuth range according to the azimuth type corresponding to each azimuth range and the target azimuth range.

Comparing a target azimuth angle range with each azimuth angle range, and determining the azimuth angle range which is the same as the target azimuth angle range in each azimuth angle range; and acquiring the azimuth type corresponding to the azimuth range which is the same as the target azimuth range in each azimuth range, and taking the azimuth type as the target azimuth type corresponding to the target azimuth range.

In one embodiment, a method for obtaining a prediction of a target pest location is provided that combines azimuth and distance to determine a prediction of a target pest location, and that is based on a spherical earth in calculating both azimuth and distance, which will have a higher degree of accuracy than a biplane-based approach. As shown in fig. 6, the obtaining a target pest location prediction result according to the pest reference location and the category center location corresponding to the target azimuth category at each of the first pest time in step 108 includes:

step 108A, calculating an azimuth angle and a distance of a category center position corresponding to the target azimuth category relative to the pest reference position at each first pest time.

The formula for calculating the azimuth angle is shown in formulas (3) to (5), and will not be described in detail here. In the embodiment of the present invention, the distance is calculated with reference to equation (6).

Wherein R is_aIs the equatorial radius of the earth, R_bDistance is the calculated class center position (longitude lon) for the polar radius of the earth_bLatitude lat_b) And insect pest reference position (longitude: lon_aLatitude lat_a) The distance of (c).

Step 108B, determining an azimuth target prediction model according to the azimuth angle of the category center position corresponding to the target azimuth category at each first pest time relative to the pest reference position, and determining a distance target prediction model according to the distance of the category center position corresponding to the target azimuth category at each first pest time relative to the pest reference position.

The model coefficients comprise variable coefficients and model constants corresponding to independent variables in the model, for example, the model y is a × x + b, x is an independent variable, y is a dependent variable, a is a variable coefficient, and b is a model constant.

In one embodiment, a method for solving model coefficients in an azimuth prediction model to obtain an azimuth target prediction model is provided. Step 108B, determining an azimuth target prediction model according to an azimuth of a category center position corresponding to the target azimuth category at each pest attack time with respect to the pest attack reference position, includes:

acquiring an azimuth angle prediction model to be solved and a first loss function, wherein the azimuth angle prediction model to be solved comprises a first coefficient to be solved; obtaining a solution of the first coefficient to be solved according to the azimuth angle of the category center position corresponding to the target azimuth category at each first pest time relative to the pest reference position and the first loss function; and substituting the solution of the first coefficient to be solved into the azimuth angle prediction model to be solved to obtain the azimuth angle target prediction model.

Wherein the azimuth angle prediction model to be solved is an azimuth angle prediction model with unknown model coefficients, for example, the azimuth angle prediction model to be solved is

x_iArranging the second pest time according to time sequence for the ith second pest time to determine x_iA value of, e.g., x_i＝[1，2，3，4，5，6，…，n]Which isWhere i is 1 … n, e.g., x for second infestation time 2000.10.01₁1, x corresponding to second infestation time 2000.10.02₂2, and so on.

And predicting the result of the azimuth corresponding to the ith second pest time.

The first loss function is a preset function for determining an optimal model coefficient in the azimuth prediction model to be solved, for example, the first loss function is:

wherein, the first coefficient to be solved is the model coefficient to be solved in the azimuth prediction model to be solved, for example,

the first coefficients to be solved are a and b.

When loss calculates a partial derivative of a, and the partial derivative is 0, the optimal solution of a is calculated; when loss calculates the partial derivative of b, and the partial derivative is 0, the optimal solution of b is obtained, as shown in the following formula.

Wherein x is_iConcentrating the ith first pest time for historical pests, likewise, x_i＝[1，2，3，4，5，6，…，n]Wherein i is 1 … n; y is_iAnd the azimuth angle of the category center position corresponding to the target azimuth category at the ith first pest attack time relative to the pest attack reference position is obtained, and n is the total number of the first pest attack time in the historical pest attack set.

After the optimal solution of a and the optimal solution of b in the first coefficient to be solved are obtained, substituting the optimal solution of a and the optimal solution of b into the azimuth angle prediction model to be solved

The azimuth angle can be obtainedAn object prediction model.

In one embodiment, a method for solving model coefficients in a distance prediction model to obtain a distance target prediction model is provided. Step 108B, determining a distance target prediction model according to a distance between a category center position corresponding to a target orientation category at each of the first pest time and the pest reference position, includes:

obtaining a distance prediction model to be solved and a second loss function, wherein the distance prediction model to be solved comprises a second coefficient to be solved; obtaining a solution of the second coefficient to be solved according to the distance between the category center position corresponding to the target azimuth category at each first pest time and the pest reference position and the second loss function; and substituting the solution of the second to-be-solved coefficient into the second to-be-solved azimuth angle prediction model to obtain the distance target prediction model.

The distance prediction model to be solved is a distance prediction model with unknown model coefficients; the second loss function is a preset function used for determining the optimal model coefficient in the distance prediction model to be solved; and the second to-be-solved coefficient is a model coefficient to be solved in the to-be-solved distance prediction model. The above-mentioned method for determining the distance target prediction model is the same as the method for determining the azimuth target prediction model, and will not be described in detail here.

And 108C, obtaining an azimuth angle prediction result according to the azimuth angle target prediction model and the second pest attack time, and obtaining a distance prediction result according to a distance target prediction model and the second pest attack time, wherein the azimuth angle prediction result is the azimuth angle of the pest position in the target azimuth angle range at the second pest attack time relative to the pest attack reference position, and the distance prediction result is the distance of the pest position in the target azimuth angle range at the second pest attack time relative to the pest attack reference position.

The azimuth angle prediction result is an azimuth angle obtained by predicting through an azimuth angle target prediction model, and specifically, x corresponding to the second pest time_iSubstituting the value into an azimuth target prediction model to obtain an azimuthPredicting the result; the distance prediction result is the distance predicted by the distance target prediction model, specifically, x corresponding to the second pest time_iSubstituting the value into the distance target prediction model to obtain a distance prediction result.

And 108D, obtaining a target insect pest position prediction result according to the azimuth angle prediction result and the distance prediction result.

Acquiring a preset azimuth angle formula and a preset distance formula, wherein the preset azimuth angle formula records the corresponding relation between an azimuth angle and a pest position, and the preset distance formula records the corresponding relation between a distance and a pest position;

and obtaining the target pest position prediction result according to the azimuth angle prediction result, the distance prediction result, the preset azimuth angle formula and the preset distance formula.

The preset azimuth angle formula can be shown as formulas (3) to (5); the preset distance formula can be shown as formula (6). Because the azimuth angle prediction result (azimuth) and the distance prediction result (distance) are obtained, the target pest position prediction result (longitude and latitude) can be obtained according to the formulas (3) to (6).

In one embodiment, environmental impact parameters are introduced to improve the accuracy of the final predicted pest location, taking into account the natural environment that may have an impact on the direction of pest development. As shown in fig. 7, the step 108D of obtaining the target pest location prediction result according to the azimuth angle prediction result and the distance prediction result includes:

and step 108D1, obtaining an environmental impact parameter corresponding to the azimuth angle prediction result and the distance prediction result, where the environmental impact parameter includes at least one of an environmental coefficient, an environmental value, and an environmental value coefficient.

The environment influence parameters are used for measuring influences of external environments on the pest position prediction result, such as wind, temperature and humidity of air and the like, and include at least one of environment coefficients, environment values and environment value coefficients, namely the environment influence parameters can only include one parameter value or can also include a plurality of parameter values.

Wherein, the environment coefficient is a preset constant value; an environment value reflecting the magnitude of the external environment (e.g., wind); the environment value coefficient is combined with the environment value to obtain the direct influence of the external environment on the azimuth angle prediction result and the distance prediction result, for example, cos (theta) for the environment value coefficient_i,v) Is represented by, wherein_i,vThe angle between the azimuth prediction and the environmental direction (e.g., wind direction) is shown in fig. 7.

And step 108D2, obtaining the target pest position prediction result according to the azimuth angle prediction result, the distance prediction result and the environment influence parameter.

When the environmental impact parameters only contain one parameter value, an updated azimuth angle prediction result is obtained according to the product of the azimuth angle prediction result and the parameter value, namely the updated azimuth angle prediction result is an azimuth angle prediction result × w, wherein w is the environmental impact parameter, an updated distance prediction result is obtained according to the product of the distance prediction result and the parameter value, and finally, a target pest position prediction result is obtained according to the updated azimuth angle prediction result and the updated distance prediction result.

When the environmental impact parameter includes a plurality of parameter values, for example, the environmental impact parameter includes an environmental coefficient, an environmental value, and an environmental value coefficient, and the updated azimuth prediction result is obtained according to the product of the azimuth prediction result and the environmental coefficient, the environmental value, and the environmental value coefficient, that is, the updated azimuth prediction result is the azimuth prediction result × w × v × cos (θ)_i,v) Where w is an environmental coefficient, v is an environmental value, cos (θ)_i,v) Similarly, an updated distance prediction result is obtained from the distance prediction result multiplied by the environment coefficient and the environment value coefficient, that is, the updated distance prediction result is × w × v × cos (θ)_i,v). And finally, obtaining a target pest position prediction result according to the updated azimuth angle prediction result and the updated distance prediction result.

In one embodiment, when the life cycle is a high emergence period, the development speed of the insect pests is extremely high, at the moment, the environmental coefficient can be set to be larger, when the life cycle is a low emergence period, the development speed of the insect pests is relatively slower, at the moment, the environmental coefficient can be set to be smaller, and therefore a more accurate prediction result is obtained. Step 108D1, the obtaining the environmental impact parameters corresponding to the azimuth angle prediction result and the distance prediction result includes:

acquiring the current life cycle of the insect pests; and determining an environment coefficient corresponding to the azimuth angle prediction result and the distance prediction result according to the current life cycle of the insect pest.

The current life cycle of the insect pests indicates the current life cycle of the insect pests, exemplarily, the life cycle comprises a primary period, a high-occurrence period and a death period, when the life cycle is the primary period, the development speed of the insect pests is very low, when the life cycle is the primary period, the insect pests start to have a certain development speed, the quantity starts to increase, when the life cycle is the high-occurrence period, the development speed of the insect pests is very high, the quantity sharply increases, and when the life cycle is the death period, the insect pests grow slowly and the quantity starts to gradually decrease.

And presetting a period coefficient table, wherein the period coefficient table records the corresponding relation between the life cycle and the candidate environment coefficient. Obtaining a candidate environment coefficient corresponding to the current life cycle of the insect pest from the cycle coefficient table; and taking the candidate environment coefficient corresponding to the current life cycle of the insect pest as the environment coefficient in the environment influence parameter.

In one embodiment, the application environment of the pest location prediction method is provided, and particularly, the pest location prediction method is applied to prediction of pests of crops in a certain industrial park or administrative area, and meanwhile, in order to guarantee effectiveness of prediction results, over-range pest location prediction results are removed. Specifically, the method further comprises: acquiring a target area identifier; acquiring a longitude and latitude range corresponding to the target area identification; and removing target pest position prediction results exceeding the latitude and longitude range from target pest position prediction results corresponding to all azimuth categories according to the latitude and longitude range corresponding to the target area identification.

The target area identifier is used for identifying a target area, and the target area may be a certain industrial park or a certain administrative area, which is not specifically limited herein; and when the target pest position prediction results corresponding to all the azimuth categories have target pest position prediction results exceeding the latitude and longitude range, removing the target pest position prediction results exceeding the latitude and longitude range, and ensuring the accuracy of the finally determined prediction results.

As shown in fig. 8, a pest location prediction device 800 is provided, which specifically includes:

a location obtaining module 802, configured to obtain a historical pest set, where the historical pest set includes a plurality of pest locations and a first pest time corresponding to each pest location;

a reference determining module 804, configured to obtain a pest reference position according to the multiple pest positions;

a category determining module 806, configured to obtain, according to each of the pest locations, a first pest time corresponding to each of the pest locations, and the pest reference location, a category center location corresponding to a target azimuth category at each of the first pest times, where the category center location is a center of a pest location within a target azimuth range, the target azimuth range is an azimuth range corresponding to the target azimuth category, the target azimuth range is any one of a plurality of preset azimuth ranges, and the plurality of azimuth ranges overlap and cover all azimuths;

and the result prediction module 808 is configured to obtain a target pest position prediction result according to the pest reference position and the category center position corresponding to the target azimuth category at each first pest time, where the target pest position prediction result is a prediction result of a pest position within the target azimuth range at a second pest time when the pest reference position is used as the reference position, and the second pest time is a time after the first pest time.

According to the insect pest position prediction device, insect pest reference positions are obtained according to a plurality of historical insect pest concentrated insect pest positions, then according to the plurality of historical insect pest concentrated insect pest positions, the first insect pest time corresponding to each insect pest position and the insect pest reference positions obtain each category center position corresponding to the target position category at the first insect pest time, and finally, according to the insect pest reference positions and each category center position corresponding to the target position category at the first insect pest time, prediction of the insect pest positions is achieved. Moreover, the prediction result of the pest position is the prediction result in the target azimuth angle range, that is, the prediction results in different azimuth angle ranges can be obtained in the above manner, and compared with the manner that only one prediction result at the second pest time can be obtained in a prediction manner, the manner can obtain a more precise prediction result.

In an embodiment, the category determining module 806 is specifically configured to: determining an azimuth angle of each pest position relative to the pest reference position according to each pest position and the pest reference position; determining pest positions belonging to the target azimuth category according to the azimuth angle of each pest position relative to the pest reference position and the target azimuth angle range; and obtaining a category central position corresponding to the target azimuth category at each first pest attack time according to the pest attack positions belonging to the target azimuth category based on the first pest attack time corresponding to the pest attack position.

In one embodiment, the apparatus 800 further comprises: the quantity category module is used for acquiring the historical pest quantity in a preset time period; determining the plurality of azimuth angle ranges according to the historical pest number in the preset time period; determining an azimuth category corresponding to each azimuth angle range; and determining a target azimuth type corresponding to the target azimuth range according to the azimuth type corresponding to each azimuth range and the target azimuth range.

In an embodiment, the result predicting module 808 is specifically configured to: calculating the azimuth angle and the distance of the category center position corresponding to the target azimuth category relative to the pest reference position at each first pest time; determining an azimuth target prediction model according to an azimuth angle of a category center position corresponding to a target azimuth category at each first pest time relative to the pest reference position, and determining a distance target prediction model according to a distance between the category center position corresponding to the target azimuth category at each first pest time relative to the pest reference position; obtaining an azimuth angle prediction result according to an azimuth angle target prediction model and the second pest attack time, and obtaining a distance prediction result according to a distance target prediction model and the second pest attack time, wherein the azimuth angle prediction result is the azimuth angle of the pest attack position in the target azimuth angle range at the second pest attack time relative to the pest attack reference position, and the distance prediction result is the distance of the pest attack position in the target azimuth angle range at the second pest attack time relative to the pest attack reference position; and obtaining the target pest position prediction result according to the azimuth angle prediction result and the distance prediction result.

In an embodiment, the result predicting module 808 is specifically configured to: acquiring an environmental influence parameter corresponding to the azimuth angle prediction result and the distance prediction result, wherein the environmental influence parameter comprises at least one of an environmental coefficient, an environmental value and an environmental value coefficient; and obtaining the target pest position prediction result according to the azimuth angle prediction result, the distance prediction result and the environment influence parameter.

In an embodiment, the result predicting module 808 is specifically configured to: acquiring the current life cycle of the insect pests; and determining an environment coefficient corresponding to the azimuth angle prediction result and the distance prediction result according to the current life cycle of the insect pest.

In an embodiment, the result predicting module 808 is specifically configured to: acquiring an azimuth angle prediction model to be solved and a first loss function, wherein the azimuth angle prediction model to be solved comprises a first coefficient to be solved; obtaining a solution of the first coefficient to be solved according to the azimuth angle of the category center position corresponding to the target azimuth category at each first pest time relative to the pest reference position and the first loss function; and substituting the solution of the first coefficient to be solved into the azimuth angle prediction model to be solved to obtain the azimuth angle target prediction model.

In an embodiment, the result predicting module 808 is specifically configured to: obtaining a distance prediction model to be solved and a second loss function, wherein the distance prediction model to be solved comprises a second coefficient to be solved; obtaining a solution of the second coefficient to be solved according to the distance between the category center position corresponding to the target azimuth category at each first pest time and the pest reference position and the second loss function; and substituting the solution of the second to-be-solved coefficient into the second to-be-solved azimuth angle prediction model to obtain the distance target prediction model.

In an embodiment, the result predicting module 808 is specifically configured to: acquiring a preset azimuth angle formula and a preset distance formula, wherein the preset azimuth angle formula records the corresponding relation between an azimuth angle and a pest position, and the preset distance formula records the corresponding relation between a distance and a pest position; and obtaining the target pest position prediction result according to the azimuth angle prediction result, the distance prediction result, the preset azimuth angle formula and the preset distance formula.

FIG. 9 is a diagram illustrating an internal structure of a computer device in one embodiment. The computer device may specifically be a terminal or a server. As shown in fig. 9, the computer device includes a processor, a memory, and a network interface connected by a system bus. Wherein the memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium of the computer device stores an operating system and may also store a computer program that, when executed by the processor, causes the processor to implement a pest location prediction method. The internal memory may also have a computer program stored therein that, when executed by the processor, causes the processor to perform a pest location prediction method. Those skilled in the art will appreciate that the architecture shown in fig. 9 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.

In one embodiment, the pest location prediction method provided herein may be implemented in the form of a computer program that is executable on a computer device such as that shown in fig. 9. The memory of the computer device may store therein respective program templates constituting the pest location predicting device. Such as a location acquisition module 802, a benchmark determination module 804, and a category determination module 806.

A computer device comprising a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of:

acquiring a historical pest set, wherein the historical pest set comprises a plurality of pest positions and a first pest time corresponding to each pest position;

obtaining pest reference positions according to the multiple pest positions;

obtaining a category central position corresponding to a target azimuth category at each first pest time according to each pest position, a first pest time corresponding to each pest position and the pest reference position, wherein the category central position is the center of the pest position within a target azimuth range, the target azimuth range is an azimuth range corresponding to the target azimuth category, the target azimuth range is any one azimuth range of a plurality of preset azimuth ranges, and the plurality of azimuth ranges are overlapped to cover all azimuths;

and obtaining a target pest position prediction result according to the pest reference position and the category central position corresponding to the target azimuth category at each first pest time, wherein the target pest position prediction result is a prediction result of a pest position within a target azimuth angle range at a second pest time when the pest reference position is used as the reference position, and the second pest time is a time after the first pest time.

In one embodiment, the obtaining of a category center position corresponding to a target azimuth category at each first pest time according to each pest location, a first pest time corresponding to each pest location, and the pest reference location includes: determining an azimuth angle of each pest position relative to the pest reference position according to each pest position and the pest reference position; determining pest positions belonging to the target azimuth category according to the azimuth angle of each pest position relative to the pest reference position and the target azimuth angle range; and obtaining a category central position corresponding to the target azimuth category at each first pest attack time according to the pest attack positions belonging to the target azimuth category based on the first pest attack time corresponding to the pest attack position.

In one embodiment, prior to said obtaining the historical set of pests, further comprising: acquiring the historical pest number in a preset time period; determining the plurality of azimuth angle ranges according to the historical pest number in the preset time period; determining an azimuth category corresponding to each azimuth angle range; and determining a target azimuth type corresponding to the target azimuth range according to the azimuth type corresponding to each azimuth range and the target azimuth range.

In one embodiment, the obtaining a target pest location prediction result according to the pest reference location and a category center location corresponding to a target azimuth category at each of the first pest time includes: calculating the azimuth angle and the distance of the category center position corresponding to the target azimuth category relative to the pest reference position at each first pest time; determining an azimuth target prediction model according to an azimuth angle of a category center position corresponding to a target azimuth category at each first pest time relative to the pest reference position, and determining a distance target prediction model according to a distance between the category center position corresponding to the target azimuth category at each first pest time relative to the pest reference position; obtaining an azimuth angle prediction result according to an azimuth angle target prediction model and the second pest attack time, and obtaining a distance prediction result according to a distance target prediction model and the second pest attack time, wherein the azimuth angle prediction result is the azimuth angle of the pest attack position in the target azimuth angle range at the second pest attack time relative to the pest attack reference position, and the distance prediction result is the distance of the pest attack position in the target azimuth angle range at the second pest attack time relative to the pest attack reference position; and obtaining the target pest position prediction result according to the azimuth angle prediction result and the distance prediction result.

In one embodiment, said obtaining said target pest location prediction based on said azimuth angle prediction and said distance prediction comprises: acquiring an environmental influence parameter corresponding to the azimuth angle prediction result and the distance prediction result, wherein the environmental influence parameter comprises at least one of an environmental coefficient, an environmental value and an environmental value coefficient; and obtaining the target pest position prediction result according to the azimuth angle prediction result, the distance prediction result and the environment influence parameter.

In one embodiment, the obtaining the environmental impact parameters corresponding to the azimuth angle prediction result and the distance prediction result includes: acquiring the current life cycle of the insect pests; and determining an environment coefficient corresponding to the azimuth angle prediction result and the distance prediction result according to the current life cycle of the insect pest.

In one embodiment, the determining an azimuth target prediction model according to an azimuth angle of a category center position corresponding to a target azimuth category at each pest time relative to the pest reference position includes: acquiring an azimuth angle prediction model to be solved and a first loss function, wherein the azimuth angle prediction model to be solved comprises a first coefficient to be solved; obtaining a solution of the first coefficient to be solved according to the azimuth angle of the category center position corresponding to the target azimuth category at each first pest time relative to the pest reference position and the first loss function; and substituting the solution of the first coefficient to be solved into the azimuth angle prediction model to be solved to obtain the azimuth angle target prediction model.

In one embodiment, determining a distance target prediction model according to the distance of the category center position corresponding to the target azimuth category at each first pest time relative to the pest reference position comprises: obtaining a distance prediction model to be solved and a second loss function, wherein the distance prediction model to be solved comprises a second coefficient to be solved; obtaining a solution of the second coefficient to be solved according to the distance between the category center position corresponding to the target azimuth category at each first pest time and the pest reference position and the second loss function; and substituting the solution of the second to-be-solved coefficient into the second to-be-solved azimuth angle prediction model to obtain the distance target prediction model.

In one embodiment, said obtaining said target pest location prediction based on said azimuth angle prediction and said distance prediction comprises: acquiring a preset azimuth angle formula and a preset distance formula, wherein the preset azimuth angle formula records the corresponding relation between an azimuth angle and a pest position, and the preset distance formula records the corresponding relation between a distance and a pest position; and obtaining the target pest position prediction result according to the azimuth angle prediction result, the distance prediction result, the preset azimuth angle formula and the preset distance formula.

In one embodiment, a computer-readable storage medium is proposed, in which a computer program is stored which, when executed by a processor, causes the processor to carry out the steps of:

obtaining pest reference positions according to the multiple pest positions;

It should be noted that the above-mentioned pest location prediction method, pest location prediction apparatus, computer device, and computer-readable storage medium belong to a general inventive concept, and the contents of the embodiments of pest location prediction method, pest location prediction apparatus, computer device, and computer-readable storage medium are mutually applicable.

Those skilled in the art will appreciate that all or a portion of the processes in the methods of the embodiments described above may be implemented by computer programs that may be stored in a non-volatile computer-readable storage medium, which when executed, may include the processes of the embodiments of the methods described above, wherein any reference to memory, storage, database or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, non-volatile memory may include read-only memory (ROM), programmable ROM (prom), electrically programmable ROM (eprom), electrically erasable programmable ROM (eeprom), or flash memory, volatile memory may include Random Access Memory (RAM) or external cache memory, RAM is available in a variety of forms, such as static RAM (sram), Dynamic RAM (DRAM), synchronous sdram (sdram), double data rate sdram (ddr sdram), enhanced sdram (sdram), synchronous link (sdram), dynamic RAM (rdram) (rdram L), direct dynamic RAM (rdram), and the like, and/or external cache memory.

The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.

The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present application. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.

Claims

1. A method for predicting a location of an insect pest, comprising:

obtaining pest reference positions according to the multiple pest positions;

2. The method of claim 1, wherein obtaining a category center location corresponding to a target orientation category at each of the first pest times based on each of the pest locations, the first pest time corresponding to each of the pest locations, and the pest reference location comprises:

determining an azimuth angle of each pest position relative to the pest reference position according to each pest position and the pest reference position;

determining pest positions belonging to the target azimuth category according to the azimuth angle of each pest position relative to the pest reference position and the target azimuth angle range;

and obtaining a category central position corresponding to the target azimuth category at each first pest attack time according to the pest attack positions belonging to the target azimuth category based on the first pest attack time corresponding to the pest attack position.

3. The method of claim 1, further comprising, prior to said obtaining a historical set of pests:

acquiring the historical pest number in a preset time period;

determining the plurality of azimuth angle ranges according to the historical pest number in the preset time period;

determining an azimuth category corresponding to each azimuth angle range;

and determining a target azimuth type corresponding to the target azimuth range according to the azimuth type corresponding to each azimuth range and the target azimuth range.

4. The method of claim 1, wherein obtaining a target pest location prediction based on the pest reference location and a category center location corresponding to a target location category at each of the first pest times comprises:

calculating the azimuth angle and the distance of the category center position corresponding to the target azimuth category relative to the pest reference position at each first pest time;

determining an azimuth target prediction model according to an azimuth angle of a category center position corresponding to a target azimuth category at each first pest time relative to the pest reference position, and determining a distance target prediction model according to a distance between the category center position corresponding to the target azimuth category at each first pest time relative to the pest reference position;

obtaining an azimuth angle prediction result according to an azimuth angle target prediction model and the second pest attack time, and obtaining a distance prediction result according to a distance target prediction model and the second pest attack time, wherein the azimuth angle prediction result is the azimuth angle of the pest attack position in the target azimuth angle range at the second pest attack time relative to the pest attack reference position, and the distance prediction result is the distance of the pest attack position in the target azimuth angle range at the second pest attack time relative to the pest attack reference position;

and obtaining the target pest position prediction result according to the azimuth angle prediction result and the distance prediction result.

5. The method of claim 4, wherein said deriving said target pest location prediction from said azimuth angle prediction and said distance prediction comprises:

acquiring an environmental influence parameter corresponding to the azimuth angle prediction result and the distance prediction result, wherein the environmental influence parameter comprises at least one of an environmental coefficient, an environmental value and an environmental value coefficient;

and obtaining the target pest position prediction result according to the azimuth angle prediction result, the distance prediction result and the environment influence parameter.

6. The method of claim 5, wherein said obtaining environmental impact parameters corresponding to said azimuth prediction and said range prediction comprises:

acquiring the current life cycle of the insect pests;

and determining an environment coefficient corresponding to the azimuth angle prediction result and the distance prediction result according to the current life cycle of the insect pest.

7. The method of any one of claims 4-6, wherein determining an azimuthal target prediction model based on an azimuthal angle of a category center location corresponding to the target orientation category at each of the pest times relative to the pest reference location comprises:

acquiring an azimuth angle prediction model to be solved and a first loss function, wherein the azimuth angle prediction model to be solved comprises a first coefficient to be solved;

obtaining a solution of the first coefficient to be solved according to the azimuth angle of the category center position corresponding to the target azimuth category at each first pest time relative to the pest reference position and the first loss function;

and substituting the solution of the first coefficient to be solved into the azimuth angle prediction model to be solved to obtain the azimuth angle target prediction model.

8. The method of any one of claims 4-6, wherein determining a distance target prediction model based on a distance of a category center location corresponding to a target orientation category from the pest reference location at each of the first pest times comprises:

obtaining a distance prediction model to be solved and a second loss function, wherein the distance prediction model to be solved comprises a second coefficient to be solved;

obtaining a solution of the second coefficient to be solved according to the distance between the category center position corresponding to the target azimuth category at each first pest time and the pest reference position and the second loss function;

and substituting the solution of the second to-be-solved coefficient into the second to-be-solved azimuth angle prediction model to obtain the distance target prediction model.

9. The method of any one of claims 4-6, wherein said deriving said target pest location prediction from said azimuth angle prediction and said distance prediction comprises:

10. A pest location prediction device, comprising:

the pest control system comprises a position acquisition module, a pest control module and a pest control module, wherein the position acquisition module is used for acquiring a historical pest set which comprises a plurality of pest positions and first pest time corresponding to each pest position;

the reference determining module is used for obtaining pest reference positions according to the multiple pest positions;

the category determination module is used for obtaining a category central position corresponding to a target azimuth category at each first pest attack time according to each pest attack position, a first pest attack time corresponding to each pest attack position and the pest attack reference position, wherein the category central position is the center of the pest attack position within a target azimuth angle range, the target azimuth angle range is an azimuth angle range corresponding to the target azimuth category, the target azimuth angle range is any one azimuth angle range of a plurality of preset azimuth angle ranges, and the plurality of azimuth angle ranges are overlapped to cover all azimuths;

and the result prediction module is used for obtaining a target pest position prediction result according to the pest reference position and the category central position corresponding to the target azimuth category at each first pest time, wherein the target pest position prediction result is a prediction result of a pest position within a target azimuth angle range at a second pest time when the pest reference position is used as the reference position, and the second pest time is a time after the first pest time.

11. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor when executing the computer program performs the steps of the pest location prediction method of any one of claims 1 to 9.

12. A computer-readable storage medium storing a computer program, wherein the computer program when executed by a processor implements the steps of the pest location prediction method of any of claims 1-9.