A risk prediction method for forest pests and diseases combined with ecological niche model

Abstract

The present invention discloses a method for predicting the risk of forestry pests and diseases combined with an ecological niche model, and relates to the technical field of data processing. The method comprises: obtaining historical pest and disease distribution data of a target forestry area within a historical observation period, determining multiple historical distribution area sets and multiple historical distribution environmental data tolerance intervals of multiple pest and disease species; obtaining multiple pest and disease time granularity curve sets; determining multiple distribution area sets to be analyzed and multiple distribution quantity sets; determining multiple pest and disease time granularity curve sets to be mapped; determining multiple environmental mapping deviation factor sets to be analyzed; performing pest and disease risk prediction, and obtaining multiple risk prediction result sets. The present invention solves the technical problem of low accuracy of forestry pest and disease risk prediction in the prior art, and achieves the technical effect of improving the reliability and efficiency of pest and disease risk prediction.

Description

Technical Field

[0001] The invention relates to the technical field of data processing, and in particular to a method for predicting forestry pest and disease disaster risks in combination with an ecological niche model. Background Art

[0002] At present, when predicting the risk of forest pests and diseases, people may only rely on single historical pest and disease data or simple statistical analysis, while ignoring the complex relationship between the occurrence of pests and diseases and various environmental factors. As a result, in practical applications, the prediction results often deviate greatly from the actual situation. Especially when facing sudden pest and disease disasters, the prediction accuracy is low. Summary of the invention

[0003] The present application provides a forestry pest and disease disaster risk prediction method combined with an ecological niche model, which is used to solve the technical problem of low accuracy in forestry pest and disease disaster risk prediction in the prior art.

[0004] In view of the above problems, the present application provides a forestry pest and disease disaster risk prediction method combined with an ecological niche model, the method comprising:

[0005] Obtain historical pest and disease distribution data of the target forestry area within the historical observation period, conduct multi-dimensional analysis on the historical pest and disease distribution data, and determine multiple historical distribution area sets and multiple historical distribution environment data tolerance intervals of multiple pest and disease species;

[0006] Based on the multiple historical distribution area sets and the historical pest and disease distribution data, construct pest and disease time granularity curves of multiple pest and disease species to obtain multiple pest and disease time granularity curve sets, wherein each pest and disease time granularity curve corresponds to a historical distribution area;

[0007] Based on the multiple historical distribution area sets, the ecological niche model of the target forestry area is retrieved to determine multiple distribution area sets to be analyzed and multiple distribution quantity sets of the multiple pest and disease species at the current moment, wherein the distribution area to be analyzed is the historical distribution area in the historical distribution area set corresponding to each pest and disease species whose distribution quantity at the current moment exceeds the preset distribution quantity;

[0008] According to the matching of the multiple distribution area sets to be analyzed with the multiple pest and disease time granularity curve sets, a plurality of pest and disease time granularity curve sets to be mapped are determined;

[0009] Interacting the ecological niche model to obtain multiple distribution environment data of the multiple distribution area sets to be analyzed at the current moment, performing environmental mapping deviation analysis in combination with the multiple historical distribution environment data tolerance intervals, and determining multiple sets of environmental mapping deviation factors to be analyzed;

[0010] Pest and disease risk prediction is performed based on the multiple distribution quantity sets, the multiple time granularity curve sets of pests and diseases to be mapped, and the multiple sets of environmental mapping deviation factors to be analyzed to obtain multiple risk prediction result sets.

[0011] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0012] The present application obtains historical pest and disease distribution data of the target forestry area within the historical observation period, performs multi-dimensional analysis on the historical pest and disease distribution data, determines multiple historical distribution area sets and multiple historical distribution environment data tolerance intervals of multiple pest and disease species, and then constructs pest and disease time granularity curves for multiple pest and disease species based on the multiple historical distribution area sets and historical pest and disease distribution data, and obtains multiple pest and disease time granularity curve sets, wherein each pest and disease time granularity curve corresponds to a historical distribution area, and then based on the multiple historical distribution area sets, retrieves the ecological niche model of the target forestry area to determine multiple distribution area sets and multiple distribution quantity sets to be analyzed for multiple pest and disease species at the current moment, wherein the distribution area to be analyzed The domain is the historical distribution area in which the distribution quantity at the current moment exceeds the preset distribution quantity in the historical distribution area set corresponding to each pest species. By matching multiple distribution area sets to be analyzed with multiple pest time granularity curve sets, multiple pest time granularity curve sets to be mapped are determined. Then, the interactive niche model obtains multiple distribution environment data of multiple distribution area sets to be analyzed at the current moment, performs environmental mapping deviation analysis in combination with multiple historical distribution environment data tolerance intervals, determines multiple environmental mapping deviation factor sets to be analyzed, and performs pest risk prediction based on multiple distribution quantity sets, multiple pest time granularity curve sets to be mapped, and multiple pest environment mapping deviation factor sets to be analyzed, and obtains multiple risk prediction result sets. The technical effect of improving the quality and reliability of risk prediction is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0014] Figure 1 A schematic diagram of a process for predicting forest pest and disease disaster risks in combination with an ecological niche model provided in an embodiment of the present application;

[0015] Figure 2A schematic diagram of a process for obtaining tolerance intervals of multiple historical distribution environment data in a forestry pest and disease disaster risk prediction method combined with an ecological niche model provided in an embodiment of the present application. DETAILED DESCRIPTION

[0016] The present application provides a forestry pest and disease disaster risk prediction method combined with an ecological niche model, which is used to solve the technical problem of low accuracy in forestry pest and disease disaster risk prediction in the prior art.

[0017] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0018] It should be noted that the terms "including" and "having" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or inherent to these processes, methods, products or devices.

[0019] Embodiment 1, as Figure 1 As shown, the present application provides a method for predicting forest pest and disease risks in combination with an ecological niche model, wherein the method comprises:

[0020] S100: Obtain historical pest and disease distribution data of the target forestry area within a historical observation period, perform multi-dimensional analysis on the historical pest and disease distribution data, and determine multiple historical distribution area sets and multiple historical distribution environment data tolerance intervals of multiple pest and disease species;

[0021] In an embodiment of the present application, the target forestry area is any area that needs to be predicted for the risk of pests and diseases, usually used for planting and managing forest resources, including shelterbelt areas, nature reserves and national parks. The historical observation period is a time period for monitoring the distribution of pests and diseases in the target forestry area in history, which can be half a year, a year, etc. The historical pest and disease distribution data is obtained by retrieving the pest and disease data of the target forestry area within the historical observation period. Among them, the historical pest and disease distribution data is used to describe the geographical distribution information, density, occurrence time, etc. of different pest and disease species in the target forestry area.

[0022] Preferably, after obtaining the historical pest and disease distribution data, the data is disassembled and analyzed from multiple dimensions, so as to extract the distribution areas and distribution environment conditions of multiple pest and disease species contained in the target forestry area during the historical observation period, and obtain multiple historical distribution area sets and multiple historical distribution environment tolerance intervals of the multiple pest and disease species. Among them, the historical distribution environment tolerance interval corresponds to a historical distribution area, which is used to describe the general environmental conditions in the area. Multiple historical distribution area sets represent the spatial distribution ranges of different pest and disease species under different time periods and environmental conditions. At the same time, based on the environmental characteristics of the pest and disease distribution area, the tolerance interval of the environmental variable is determined, that is, the range of environmental conditions (such as temperature range, humidity range, etc.) that the species can adapt to in the historical distribution area.

[0023] When predicting the risk of pests and diseases occurring in the target forestry area by different pest and disease species, it is necessary to consider the distribution areas of the different pest and disease species and the regional environment in which they are distributed, so as to provide reliable data support for subsequent pest and disease disaster predictions tailored to the actual conditions of the target forestry area.

[0024] Further, historical pest distribution data of the target forestry area within the historical observation period is obtained, and the historical pest distribution data is subjected to multi-dimensional analysis to determine multiple historical distribution area sets and multiple historical distribution environment tolerance intervals of multiple pest species. Step S100 of the embodiment of the present application further includes:

[0025] Using pest and disease species as an index, searching the historical pest and disease distribution data to determine a plurality of initial historical distribution area sets, wherein the plurality of initial historical distribution area sets include a plurality of regional distribution location sets;

[0026] Performing nearest neighbor fusion on the multiple initial historical distribution area sets based on the multiple area distribution location sets to determine multiple historical distribution area sets;

[0027] Taking the multiple historical distribution area sets as analysis objects, environmental monitoring data are retrieved to narrow the scope of environmental data, and multiple tolerance intervals of historical distribution environmental data are determined.

[0028] In a possible embodiment, the distribution records of pests and diseases species are retrieved based on the data collected in the early stage, and multiple initial historical distribution area sets of each species are preliminarily determined. The multiple initial historical distribution area sets reflect the historical distribution of species in the target forestry area. Based on the multiple regional distribution position sets, the multiple initial historical distribution area sets are subjected to two-level neighbor area analysis, and the multiple initial historical distribution area sets are fused according to the analysis results, thereby improving the analysis efficiency on the basis of ensuring the distribution reliability of different pests and diseases species.

[0029] Optionally, after determining multiple sets of historical distribution areas, the environmental conditions in different historical distribution areas are effectively limited in combination with the environmental monitoring data of the target forestry area, and the environmental monitoring data that appears occasionally or with low frequency is eliminated to improve the accuracy of risk prediction, thereby obtaining the tolerance intervals of the multiple historical distribution environmental data. By obtaining multiple tolerance intervals of historical distribution environmental data, it can help identify the potential distribution areas of different pests and diseases in the future or under different environmental conditions, as well as the environmental conditions of pests and diseases in different historical distribution areas, and also provide data support for predicting future pest and disease risks.

[0030] Further, based on the multiple regional distribution position sets, the multiple initial historical distribution area sets are subjected to nearest neighbor fusion to determine multiple historical distribution area sets. Step S100 of the embodiment of the present application further includes:

[0031] Selecting a plurality of central initial historical distribution area sets from the plurality of initial historical distribution area sets respectively, and performing first-level adjacent area screening on the plurality of central initial historical distribution area sets in combination with the plurality of area distribution position sets to obtain a plurality of first-level area neighborhood sets, wherein a first-level area neighborhood is an initial historical distribution area set, and the plurality of first-level area neighborhood sets include a plurality of first-level area spacing clusters;

[0032] Performing secondary adjacent neighborhood screening on the multiple first-level regional neighborhood sets in combination with the multiple regional distribution position sets to obtain multiple second-level regional neighborhood sets, wherein the second-level regional neighborhood is an initial historical distribution area in the multiple initial historical distribution area sets that is adjacent to multiple initial historical distribution areas in a regional neighborhood, and the multiple second-level regional neighborhood sets include multiple second-level regional spacing clusters;

[0033] A plurality of historical distribution area sets are determined by performing neighbor fusion based on the plurality of first-level area neighborhood sets, the plurality of first-level area spacing clusters, the plurality of second-level area neighborhood sets, and the plurality of second-level area spacing clusters.

[0034] Further, according to the multiple first-level area neighborhood sets, the multiple first-level area spacing clusters, the multiple second-level area neighborhood sets, and the multiple second-level area spacing clusters, a plurality of historical distribution area sets are determined. Step S100 of the embodiment of the present application further includes:

[0035] Performing weighted calculation on the plurality of primary region spacing clusters and the plurality of secondary region spacing clusters according to preset weights to determine a plurality of weighted spacing clusters;

[0036] It is determined whether the plurality of weighted distance clusters are greater than or equal to a preset distance threshold; if not, a historical distribution area is added to the corresponding primary area neighborhood and secondary area neighborhood to obtain the plurality of historical distribution area sets.

[0037] In a possible embodiment, multiple central initial historical distribution area sets are respectively selected from the multiple initial historical distribution area sets, wherein the number of areas included in each central initial historical distribution area set is set by a person skilled in the art according to the number of areas included in the central initial historical area. Preferably, the distance between any two central initial historical distribution areas in each central initial historical distribution area set is greater than or equal to a preset distance, thereby ensuring that the selected multiple central initial historical distribution area sets can represent the distribution of the multiple pest and disease species.

[0038] Furthermore, by calculating the geographical distance between each initial historical distribution area set and multiple central initial historical distribution areas, adjacent areas are identified. An initial historical distribution area whose geographical distance to multiple central initial historical distribution areas is less than a spacing threshold pre-set by a person skilled in the art is defined as a first-level adjacent area. Each first-level area neighborhood set represents an adjacent initial historical distribution area of ​​a central initial historical distribution area.

[0039] Furthermore, according to the distance between any initial historical distribution area in the multiple first-level regional neighborhood sets and a central initial historical distribution area, multiple first-level regional distance clusters are obtained. Based on the first-level regional neighborhood set, other initial historical distribution areas adjacent to the initial historical distribution areas in the first-level regional neighborhood set are further screened and defined as second-level adjacent neighborhoods. Similar to the first-level regional distance cluster, the geographical distance of the second-level adjacent areas is calculated and divided into multiple second-level regional distance clusters.

[0040] By integrating adjacent initial historical distribution areas into a larger set of areas, the spatial distribution pattern of pests and diseases is reflected. Preferably, multiple weighted spacing clusters are determined by weighted calculation of the primary regional spacing clusters and the secondary regional spacing clusters according to preset weights. The weight setting can be adjusted based on factors such as regional proximity, spatial distance or environmental similarity.

[0041] Determine whether the weighted distance in each weighted distance cluster is greater than or equal to a preset distance threshold. If the value of the weighted distance cluster is less than the preset threshold, it means that the geographical distance between adjacent areas is small and the distribution is relatively close, then the corresponding first-level area neighborhood and second-level area neighborhood are merged as a new historical distribution area. This step ensures that adjacent and closely distributed areas can be integrated into a larger area set. Thus, the multiple historical distribution area sets are obtained.

[0042] Through nearest neighbor fusion, multiple initial historical distribution area sets are screened, and then by constructing primary and secondary adjacent neighborhood sets, calculating weighted interval clusters and performing threshold judgment, multiple historical distribution area sets can be finally determined. These area sets provide data support for subsequent pest and disease risk prediction and help determine the spatial distribution characteristics of pests and diseases and their response to environmental changes.

[0043] Further, taking the multiple historical distribution area sets as analysis objects, retrieving environmental monitoring data to narrow the environmental data range, and determining multiple historical distribution environmental data tolerance intervals, step S100 of the embodiment of the present application includes:

[0044] Taking the multiple historical distribution area sets as analysis objects, retrieving environmental monitoring data of each historical distribution area to obtain multiple environmental monitoring data sets;

[0045] Traversing the plurality of environmental monitoring data sets to perform mean calculations and determine a plurality of environmental monitoring average data;

[0046] The plurality of environmental monitoring average data are used as a plurality of narrowing centers, and data ranges of the plurality of environmental monitoring data sets are narrowed to determine tolerance intervals of the plurality of historical distribution environmental data.

[0047] Further, such as Figure 2 As shown, step S100 in the embodiment of the present application also includes:

[0048] Taking the multiple narrowing centers as regional centers, and combining a preset search step length with the multiple environmental monitoring data sets to construct multiple initial data tolerance interval sets;

[0049] Perform edge diffusion on the multiple initial data tolerance interval sets respectively according to a preset search step length to obtain multiple diffusion data tolerance interval sets;

[0050] Determine whether the data concentration of the multiple diffusion data tolerance interval sets is greater than or equal to the data concentration of the multiple initial data tolerance intervals. If so, continue to perform edge diffusion on the multiple diffusion data tolerance interval sets according to the preset search step until the double stop condition is met, and use the multiple diffusion data tolerance intervals obtained from the last diffusion as the multiple historical distribution environment data tolerance intervals.

[0051] Furthermore, the dual stop condition is to stop edge diffusion when the data concentration of the current diffusion data tolerance interval is less than the data concentration of the previous diffusion data tolerance interval or when the data concentration of the current diffusion data tolerance interval is greater than the data concentration of the previous diffusion data tolerance interval and the data concentration difference between two adjacent diffusion data tolerance intervals is less than or equal to a preset data concentration difference.

[0052] In one embodiment, the environmental monitoring sensors of the target forestry area within a preset historical observation period are interacted, and the environmental monitoring data of each historical distribution area are retrieved with the multiple historical distribution area sets as the analysis object, so as to obtain the multiple environmental monitoring data sets. Among them, the multiple environmental monitoring data sets reflect the regional environmental conditions when disasters occur in the multiple pest and disease species within the multiple historical distribution area sets. In order to improve the reliability of subsequent prediction of pest and disease disaster risks, it is necessary to limit the data of the multiple environmental monitoring data sets, that is, to narrow the scope of the environmental monitoring data to the interval where the region frequently appears, so as to obtain the tolerance intervals of the multiple historical distribution environmental data.

[0053] In one embodiment, the multiple environmental monitoring data sets are traversed to perform mean calculation to determine the average environmental conditions of multiple historical distribution area sets. Further, the multiple environmental monitoring average data are used as multiple narrowing centers, and a historical distribution environmental data tolerance interval is constructed for each narrowing center in combination with a preset search step and multiple environmental monitoring data sets to obtain the multiple historical distribution environmental data tolerance intervals.

[0054] Preferably, the preset search step is a data difference value for a single interval diffusion pre-set by a person skilled in the art. With the multiple narrowing centers as the regional centers and the preset search step as the data difference value from the two end points of the interval to the regional center, data search is performed in the multiple environmental monitoring data sets to determine multiple initial data tolerance interval sets.

[0055] Furthermore, the preset search step is used as the data difference of the two endpoints of the initial data tolerance interval moving toward both ends respectively, so as to achieve the goal of edge diffusion of the multiple initial data tolerance interval sets respectively, thereby obtaining multiple diffusion data tolerance interval sets.

[0056] Preferably, the amount of data in the diffusion data tolerance interval is counted, and the statistical result is compared with the difference between the left and right end points of the interval to obtain data concentration. The data concentration reflects the degree of data concentration in each diffusion data tolerance interval. Based on the above principle of obtaining data concentration, the data concentration of the plurality of diffusion data tolerance interval sets and the data concentration of the plurality of initial data tolerance intervals are calculated respectively.

[0057] Furthermore, it is determined whether the data concentration of the multiple diffusion data tolerance interval sets is greater than or equal to the data concentration of the multiple initial data tolerance intervals. If so, edge diffusion is continued for the multiple diffusion data tolerance interval sets according to the preset search step until the double stop condition is met, and the multiple diffusion data tolerance intervals obtained by the last diffusion are used as multiple historical distribution environment data tolerance intervals.

[0058] Furthermore, the double stop condition is to stop edge diffusion when the data concentration of the current diffusion data tolerance interval is less than the data concentration of the previous diffusion data tolerance interval or when the data concentration of the current diffusion data tolerance interval is greater than the data concentration of the previous diffusion data tolerance interval and the data concentration difference between two adjacent diffusion data tolerance intervals is less than or equal to the preset data concentration difference. The technical effect of providing an important basis for subsequent risk prediction is achieved.

[0059] S200: constructing pest time granularity curves of multiple pest species based on the multiple historical distribution area sets and the historical pest distribution data to obtain multiple pest time granularity curve sets, wherein each pest time granularity curve corresponds to a historical distribution area;

[0060] Further, based on the multiple historical distribution area sets and the historical pest distribution data, pest time granularity curves of multiple pest species are constructed to obtain multiple pest time granularity curve sets. Step S200 of the embodiment of the present application further includes:

[0061] A time-granularity coordinate system is constructed with time as the horizontal axis and the number of pest and disease species distribution as the vertical axis;

[0062] Using the multiple historical distribution area sets as indexes, searching the historical pest and disease distribution data to obtain multiple regional pest and disease distribution data sets;

[0063] The plurality of regional pest and disease distribution data sets are respectively input into the time granularity coordinate system to obtain the plurality of pest and disease time granularity curve sets.

[0064] In one possible embodiment, a coordinate system is first created for drawing a time granularity curve of pests and diseases, wherein the time granularity curve of pests and diseases is used to describe the distribution quantity of a pest and disease species in different historical distribution areas within the target forestry area over time. Optionally, first determine the time granularity, such as dividing time by year, month, day or hour, with time as the horizontal axis (X-axis), usually arranged from left to right in chronological order. Then, the number of pest and disease species distribution is used as the vertical axis (Y-axis) to represent the number of pest and disease species at different time points. Set the scale, label and range of the coordinate system to ensure that the data can be presented clearly.

[0065] In one embodiment, multiple historical distribution area sets are used as indexes. Pest distribution data related to these areas are retrieved from the historical pest distribution database. The retrieved data are organized into multiple regional pest distribution data sets by region, each set corresponding to a historical distribution area. The regional pest distribution data are input into a time granularity coordinate system to construct a pest time granularity curve, thereby obtaining the multiple pest time granularity curve sets.

[0066] In one embodiment, for each regional pest distribution data set, the data points therein are traversed. The time information of each data point is mapped to the X-axis, and the number of pest species is mapped to the Y-axis. The data points are plotted in the coordinate system and connected into lines to form a pest time granularity curve. The above process is repeated to draw a pest time granularity curve for each historical distribution area, and multiple pest time granularity curve sets are obtained. Preferably, each curve is checked to see if it is smooth and continuous to avoid abnormal fluctuations or breakpoints.

[0067] Through step 200, a set of time granularity curves of multiple pest species can be constructed. These curves can intuitively show the distribution trend of pest species in different historical distribution areas over time, providing an important basis for subsequent pest prediction and prevention and control.

[0068] S300: Based on the multiple historical distribution area sets, the ecological niche model of the target forestry area is retrieved to determine multiple distribution area sets to be analyzed and multiple distribution quantity sets of the multiple pest and disease species at the current moment, wherein the distribution area to be analyzed is the historical distribution area in the historical distribution area set corresponding to each pest and disease species whose distribution quantity at the current moment exceeds the preset distribution quantity;

[0069] In one possible embodiment, the ecological niche model is generally used to describe the living conditions, distribution range and relationship with other species of pests and diseases species in the target forestry area. Interactive ecological niche model, obtains pest and disease species distribution data at the current moment. Preferably, the historical distribution area set corresponding to each pest and disease species is used as an index to search in the ecological niche model, thereby obtaining multiple distribution areas to be analyzed and multiple distribution quantity sets of multiple pest and disease species at the current moment. The distribution area to be analyzed is the historical distribution area in the historical distribution area set corresponding to each pest and disease species whose distribution quantity at the current moment exceeds the preset distribution quantity;

[0070] For example, by using the niche model, combined with the current environmental data of the target forestry area (such as climate, soil, vegetation, etc.), the distribution of each pest species at the current moment can be predicted or estimated. This has achieved the technical effect of providing an important reference for subsequent pest risk prediction and prevention and control.

[0071] S400: Determine a plurality of time granularity curve sets of pests and diseases to be mapped by matching the plurality of distribution area sets to be analyzed with the plurality of time granularity curve sets of pests and diseases;

[0072] S500: Interacting with the ecological niche model to obtain a plurality of distribution environment data of the plurality of distribution area sets to be analyzed at the current moment, performing environmental mapping deviation analysis in combination with the plurality of tolerance intervals of the historical distribution environment data, and determining a plurality of environmental mapping deviation factor sets to be analyzed;

[0073] In one embodiment, since each pest time granularity interval corresponds to a distribution area to be analyzed, based on the mapping relationship between the two, the plurality of pest time granularity curve sets are used as indexes to search and determine the corresponding plurality of pest time granularity curve sets to be mapped. The plurality of pest time granularity curve sets to be mapped reflect the distribution quantity of pest species in the plurality of distribution area sets to be analyzed over time.

[0074] In one embodiment, the coordinates or range of the set of distribution areas to be analyzed are input through the interactive interface with the niche model. The relevant functions of the niche model are called to obtain the climate, soil, vegetation, terrain and other environmental data of these areas at the current moment. Ensure that the acquired data is accurate, complete, and meets the analysis requirements. Compare the current distribution environment data with the tolerance interval of the historical distribution environment data to perform environmental mapping deviation analysis.

[0075] Preferably, by calculating the difference between the distribution environment data and the tolerance interval of the historical distribution environment data, and comparing the difference with the tolerance interval of the corresponding historical distribution environment data, the corresponding environment mapping deviation factor to be analyzed is obtained. The environment mapping deviation factor to be analyzed reflects the degree of deviation between the environment of the set of multiple distribution areas to be analyzed and the general environment where the corresponding pest species live.

[0076] Through step S500, the interactive niche model can obtain the distribution environment data of multiple distribution area sets to be analyzed at the current moment, and combine the tolerance interval of historical distribution environment data to perform environmental mapping deviation analysis, and finally determine multiple sets of environmental mapping deviation factors to be analyzed. These factors can provide important references for subsequent pest control and management, and help formulate targeted prevention and control strategies and measures.

[0077] S600: performing pest and disease risk prediction according to the multiple distribution quantity sets, the multiple pest and disease time granularity curve sets to be mapped, and the multiple environmental mapping deviation factor sets to be analyzed to obtain multiple risk prediction result sets.

[0078] In one embodiment, a pest and disease risk predictor is constructed, and the multiple distribution quantity sets, the multiple pest and disease time granularity curve sets to be mapped, and the multiple environmental mapping deviation factor sets to be analyzed are input into the pest and disease risk predictor for intelligent analysis and prediction to obtain the multiple risk prediction result sets.

[0079] Preferably, a plurality of sample distribution quantity sets, a plurality of sample pest and disease time granularity curve sets to be mapped, a plurality of sample environmental mapping deviation factor sets to be analyzed, and a plurality of corresponding sample pest and disease risk results are obtained as training samples. The training samples are used to perform supervised training on a network framework constructed based on a feedforward neural network, and the many-to-one mapping relationship between the distribution quantity, the pest and disease time granularity curve to be mapped, and the plurality of environmental mapping deviation factors to be analyzed and the pest and disease risk results is learned during the training until the loss function reaches convergence, and the pest and disease risk predictor that has been trained is obtained. The technical effect of improving the accuracy and reliability of forest pest and disease disaster risk prediction is achieved.

[0080] In summary, the embodiments of the present application have at least the following technical effects:

[0081] Through multidimensional analysis and fine-grained time-granularity curve construction, this application can more accurately capture the distribution and dynamic changes of pests and diseases, and combined with niche models and environmental mapping deviation analysis, it can timely identify potential high-risk areas and time periods, thereby improving the accuracy of risk prediction.

[0082] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0083] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0084] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A forestry pest and disease risk prediction method based on an ecological niche model, characterized in that: The method comprises: Obtain historical pest and disease distribution data of the target forestry area within the historical observation period, conduct multi-dimensional analysis on the historical pest and disease distribution data, and determine multiple historical distribution area sets and multiple historical distribution environment data tolerance intervals of multiple pest and disease species; Based on the multiple historical distribution area sets and the historical pest and disease distribution data, construct pest and disease time granularity curves of multiple pest and disease species to obtain multiple pest and disease time granularity curve sets, wherein each pest and disease time granularity curve corresponds to a historical distribution area; Based on the multiple historical distribution area sets, the ecological niche model of the target forestry area is retrieved to determine multiple distribution area sets to be analyzed and multiple distribution quantity sets of the multiple pest and disease species at the current moment, wherein the distribution area to be analyzed is the historical distribution area in the historical distribution area set corresponding to each pest and disease species whose distribution quantity at the current moment exceeds the preset distribution quantity; According to the matching of the plurality of distribution area sets to be analyzed with the plurality of pest and disease time granularity curve sets, a plurality of pest and disease time granularity curve sets to be mapped are determined; Interacting the ecological niche model to obtain multiple distribution environment data of the multiple distribution area sets to be analyzed at the current moment, performing environmental mapping deviation analysis in combination with the multiple historical distribution environment data tolerance intervals, and determining multiple sets of environmental mapping deviation factors to be analyzed; Pest and disease risk prediction is performed based on the multiple distribution quantity sets, the multiple time granularity curve sets of pests and diseases to be mapped, and the multiple sets of environmental mapping deviation factors to be analyzed to obtain multiple risk prediction result sets.

2. The forestry pest and disease disaster risk prediction method combined with the ecological niche model according to claim 1, characterized in that: Obtain historical pest and disease distribution data of the target forestry area during the historical observation period, conduct multi-dimensional analysis on the historical pest and disease distribution data, and determine multiple historical distribution area sets and multiple historical distribution environment tolerance intervals of multiple pest and disease species, including: Using pest and disease species as an index, searching the historical pest and disease distribution data to determine a plurality of initial historical distribution area sets, wherein the plurality of initial historical distribution area sets include a plurality of regional distribution location sets; Performing nearest neighbor fusion on the multiple initial historical distribution area sets based on the multiple area distribution location sets to determine multiple historical distribution area sets; Taking the multiple historical distribution area sets as analysis objects, environmental monitoring data are retrieved to narrow the scope of environmental data, and multiple tolerance intervals of historical distribution environmental data are determined.

3. The forestry pest and disease disaster risk prediction method combined with the ecological niche model as claimed in claim 2 is characterized in that: Performing nearest neighbor fusion on the multiple initial historical distribution area sets based on the multiple area distribution location sets to determine multiple historical distribution area sets includes: Selecting a plurality of central initial historical distribution area sets from the plurality of initial historical distribution area sets respectively, and performing first-level adjacent area screening on the plurality of central initial historical distribution area sets in combination with the plurality of area distribution position sets to obtain a plurality of first-level area neighborhood sets, wherein a first-level area neighborhood is an initial historical distribution area set, and the plurality of first-level area neighborhood sets include a plurality of first-level area spacing clusters; Performing secondary adjacent neighborhood screening on the multiple first-level regional neighborhood sets in combination with the multiple regional distribution position sets to obtain multiple second-level regional neighborhood sets, wherein the second-level regional neighborhood is an initial historical distribution area in the multiple initial historical distribution area sets that is adjacent to multiple initial historical distribution areas in a regional neighborhood, and the multiple second-level regional neighborhood sets include multiple second-level regional spacing clusters; A plurality of historical distribution area sets are determined by performing neighbor fusion based on the plurality of first-level area neighborhood sets, the plurality of first-level area spacing clusters, the plurality of second-level area neighborhood sets, and the plurality of second-level area spacing clusters.

4. The forestry pest and disease disaster risk prediction method combined with the ecological niche model as claimed in claim 3 is characterized in that: Performing neighbor fusion according to the multiple first-level area neighborhood sets, the multiple first-level area spacing clusters, the multiple second-level area neighborhood sets, and the multiple second-level area spacing clusters to determine multiple historical distribution area sets, including: Performing weighted calculation on the plurality of primary region spacing clusters and the plurality of secondary region spacing clusters according to preset weights to determine a plurality of weighted spacing clusters; It is determined whether the plurality of weighted distance clusters are greater than or equal to a preset distance threshold; if not, a historical distribution area is added to the corresponding primary area neighborhood and secondary area neighborhood to obtain the plurality of historical distribution area sets.

5. The forestry pest and disease disaster risk prediction method combined with the ecological niche model as claimed in claim 2, characterized in that: Taking the multiple historical distribution area sets as the analysis object, retrieving environmental monitoring data to narrow the environmental data range, and determining multiple historical distribution environmental data tolerance intervals, including: Taking the multiple historical distribution area sets as analysis objects, retrieving environmental monitoring data of each historical distribution area to obtain multiple environmental monitoring data sets; Traversing the plurality of environmental monitoring data sets to perform mean calculations and determine a plurality of environmental monitoring average data; The plurality of environmental monitoring average data are used as a plurality of narrowing centers, and data ranges of the plurality of environmental monitoring data sets are narrowed to determine tolerance intervals of the plurality of historical distribution environmental data.

6. The forestry pest and disease disaster risk prediction method combined with the ecological niche model as claimed in claim 5, characterized in that: include: Taking the multiple narrowing centers as regional centers, and combining a preset search step length with the multiple environmental monitoring data sets to construct multiple initial data tolerance interval sets; Perform edge diffusion on the multiple initial data tolerance interval sets respectively according to a preset search step length to obtain multiple diffusion data tolerance interval sets; Determine whether the data concentration of the multiple diffusion data tolerance interval sets is greater than or equal to the data concentration of the multiple initial data tolerance intervals. If so, continue to perform edge diffusion on the multiple diffusion data tolerance interval sets according to the preset search step until the double stop condition is met, and use the multiple diffusion data tolerance intervals obtained from the last diffusion as the multiple historical distribution environment data tolerance intervals.

7. The forestry pest and disease disaster risk prediction method combined with the ecological niche model according to claim 6, characterized in that: The dual stop condition is to stop edge diffusion when the data concentration of the current diffusion data tolerance interval is less than the data concentration of the previous diffusion data tolerance interval or when the data concentration of the current diffusion data tolerance interval is greater than the data concentration of the previous diffusion data tolerance interval and the data concentration difference between two adjacent diffusion data tolerance intervals is less than or equal to a preset data concentration difference.

8. The forestry pest and disease disaster risk prediction method combined with the ecological niche model according to claim 1, characterized in that: Based on the multiple historical distribution area sets and the historical pest distribution data, pest time granularity curves of multiple pest species are constructed to obtain multiple pest time granularity curve sets, including: A time-granularity coordinate system is constructed with time as the horizontal axis and the number of pest and disease species distribution as the vertical axis; Using the multiple historical distribution area sets as indexes, searching the historical pest and disease distribution data to obtain multiple regional pest and disease distribution data sets; The plurality of regional pest and disease distribution data sets are respectively input into the time granularity coordinate system to obtain the plurality of pest and disease time granularity curve sets.

Images