Intelligent agricultural disease and pest monitoring method and system

Abstract

The invention discloses an intelligent agricultural disease and pest monitoring method and system. The method comprises the steps of data acquisition, data preprocessing, disease and pest risk assessment model construction and agricultural disease and pest monitoring. The invention belongs to the field of agricultural monitoring, and particularly relates to an intelligent agricultural disease and pest monitoring method and system.According to the scheme, environment noise suppression measurement is introduced, abnormal monitoring data are recognized and removed, and a time attenuation factor and the Euclidean distance between samples are introduced, so that noise suppression not only considers the similarity between the samples, but also considers the similarity between the samples; dynamic changes of time and space are also considered; the agricultural disease and pest monitoring precision and the real-time processing capability are improved; a disease and pest risk assessment value is optimized by defining a dynamic adjustment function, the assessment value is dynamically adjusted according to prediction errors of different sample data, and the prediction precision is improved; by designing disease and insect pest dynamic trend loss and introducing retrospective residual optimization, errors are continuously reduced in multiple predictions, and the agricultural disease and insect pest monitoring effect is further improved.

Description

Technical Field

[0001] The present invention relates to the field of agricultural monitoring technology, and in particular to a smart agricultural pest and disease monitoring method and system. Background Art

[0002] The agricultural pest and disease monitoring method is to timely identify, predict and evaluate the pest and disease risks that crops may face by collecting and analyzing data related to crop pests and diseases, so as to take appropriate prevention and control measures and reduce the impact of pests and diseases on agricultural production. However, the general agricultural pest and disease monitoring method has the problem of poor accuracy of agricultural pest and disease monitoring due to noise, environmental changes, equipment errors and failures; the general agricultural pest and disease monitoring method has the problem of ignoring errors and uncertainties, and the risk assessment of future pests and diseases is inaccurate or too rough when facing high volatility and complex environments, which leads to poor pest and disease monitoring results. Summary of the invention

[0003] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides a smart agricultural pest and disease monitoring method and system. In view of the problem that the general agricultural pest and disease monitoring method is affected by noise, environmental changes, equipment errors and failures, which leads to poor accuracy of agricultural pest and disease monitoring, this scheme introduces environmental noise suppression metrics, identifies and removes abnormal monitoring data, introduces time attenuation factors and Euclidean distances between samples, so that noise suppression not only considers the similarity between samples, but also considers the dynamic changes of time and space; improves the accuracy and real-time processing capabilities of agricultural pest and disease monitoring; in view of the problem that the general agricultural pest and disease monitoring method ignores errors and uncertainties, and the risk assessment of future pests and diseases is inaccurate or too rough when facing high volatility and complex environments, which leads to poor pest and disease monitoring effects, this scheme optimizes the pest and disease risk assessment value by defining a dynamic adjustment function, dynamically adjusts the assessment value according to the prediction error of different sample data, and improves the prediction accuracy; by designing the dynamic trend loss of pests and diseases, and introducing retrospective residual optimization, the error is continuously reduced in multiple predictions, thereby improving the effect of agricultural pest and disease monitoring.

[0004] The technical solution adopted by the present invention is as follows: The present invention provides a smart agricultural pest monitoring method, the method comprising the following steps:

[0005] Step S1: data collection;

[0006] Step S2: data preprocessing;

[0007] Step S3: constructing a pest and disease risk assessment model;

[0008] Step S4: Agricultural pest and disease monitoring.

[0009] Furthermore, in step S1, the data collection is to collect historical agricultural pest and disease monitoring data; the historical agricultural pest and disease monitoring data includes pest and disease type data, time, geographic space data, climate and environment data, crop growth data, pesticide data and pest and disease risk assessment results; the pest and disease risk assessment results include normal, mild danger, moderate danger and severe danger; the pest and disease risk assessment results are used as data labels.

[0010] Furthermore, in step S2, the data preprocessing is to clean, convert, standardize and optimize the collected data to construct a time series data set; the data cleaning is to process missing values ​​and duplicate values; the data conversion is to convert the collected data into a vector form; the standardization is to standardize the data based on the maximum and minimum normalization method; the data optimization is to delete the data if the environmental noise suppression metric value of the data and the other 90% of the data is lower than the metric threshold; the environmental noise suppression metric value between the data is It is expressed as: ; Where A and B are different sample data; is the normalization coefficient; N is the total number of samples; and They are shape parameter and scale parameter respectively; is the gamma function; is the time decay factor; is the sample sampling time; t is the current time; is the weight decay coefficient; d(·) is the Euclidean distance; is the current sample; is the average sample of the dataset.

[0011] Furthermore, in step S3, the construction of the pest risk assessment model specifically includes the following steps:

[0012] Step S31: Preliminary construction: predict the pest risk assessment value based on LSTM, and optimize the pest risk assessment value by defining a dynamic adjustment function, thereby constructing a pest risk assessment model; define a dynamic adjustment function , expressed as: ; ; The updated formula for pest and disease risk assessment after dynamic adjustment function optimization is: ;in, is the sample data; is the prediction error of LSTM for sample data; x is the predicted pest risk value of LSTM; and are the pest and disease risk assessment values ​​before and after optimization, respectively; is the adaptive weight function; is the noise threshold; is the uncertainty of the i-th sample; is the weight decay coefficient; is the scale factor;

[0013] Step S32: Design the dynamic trend loss of pests and diseases; the dynamic trend loss ML of pests and diseases is expressed as: ; Where N is the total sample size; and are the actual pest risk value and the predicted actual pest risk value of the i-th sample respectively; and are the actual pest risk value and the predicted actual pest risk value of the i-1th sample respectively; is a smoothing term; is the trend mutation penalty factor; is the risk change;

[0014] Step S33: Review residual optimization; review residual function It is expressed as: ;in, is the prediction error of the i-th sample; is the threshold used to control the penalty residual; and Used to adjust nonlinear shapes; is the residual constraint factor; m is the number of samples reviewed; j is the sample index; is the prediction error of the ijth sample; Q is used to control the rate of exponential decay;

[0015] Step S34: construct the final loss function; the loss function TL of the pest risk assessment model is expressed as: ;in, is the loss weight;

[0016] Step S35: model determination; the time series data set corresponding to the historical agricultural pest and disease monitoring data is divided into a test set and a training set in advance; the pest and disease risk assessment model uses a gradient descent algorithm to update parameters; when the pest and disease risk assessment model converges with the training set loss, the pest and disease risk assessment model training is completed; a prediction threshold is set in advance, and when the prediction accuracy of the trained pest and disease risk assessment model for the test set is higher than the prediction threshold, the pest and disease risk assessment model is established; otherwise, the data set is re-divided, the initial parameters are adjusted, and re-training is performed.

[0017] Furthermore, in step S4, the agricultural pest and disease monitoring is based on the established pest and disease risk assessment model, and the agricultural pest and disease monitoring data is collected in real time, and the agricultural pest and disease monitoring is realized based on the pest and disease risk assessment results predicted by the model.

[0018] The invention provides a smart agricultural pest monitoring system, comprising a data acquisition module, a data preprocessing module, a pest risk assessment model building module and an agricultural pest monitoring module;

[0019] The data acquisition module collects historical agricultural pest monitoring data and sends the data to the data preprocessing module;

[0020] The data preprocessing module performs data cleaning, data conversion, standardization and data optimization on the collected data to construct a time series data set; and sends the data to the pest risk assessment model construction module;

[0021] The pest risk assessment model building module defines a dynamic adjustment function, designs the pest dynamic trend loss and introduces retrospective residual optimization, thereby building a pest risk assessment model; and sends the data to the agricultural pest monitoring module;

[0022] The agricultural pest and disease monitoring module realizes agricultural pest and disease monitoring based on the agricultural pest and disease monitoring data collected in real time based on the established pest and disease risk assessment model.

[0023] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0024] (1) In view of the problem that general agricultural pest and disease monitoring methods are affected by noise, environmental changes, equipment errors and failures, which leads to poor accuracy in agricultural pest and disease monitoring, this scheme introduces environmental noise suppression metrics to identify and remove abnormal monitoring data, and introduces a time attenuation factor and the Euclidean distance between samples, so that noise suppression not only considers the similarity between samples, but also the dynamic changes in time and space, thereby improving the accuracy and real-time processing capabilities of agricultural pest and disease monitoring.

[0025] (2) In view of the problem that general agricultural pest and disease monitoring methods ignore errors and uncertainties, and the risk assessment of future pests and diseases is inaccurate or too rough when facing high volatility and complex environments, which leads to poor pest and disease monitoring results, this scheme optimizes the pest and disease risk assessment value by defining a dynamic adjustment function, dynamically adjusts the assessment value according to the prediction error of different sample data, and improves the prediction accuracy; by designing the dynamic trend loss of pests and diseases and introducing retrospective residual optimization, the error is continuously reduced in multiple predictions, thereby improving the agricultural pest and disease monitoring effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 A schematic diagram of a process flow of a smart agricultural pest monitoring method provided by the present invention;

[0027] Figure 2 A schematic diagram of a smart agricultural pest monitoring system provided by the present invention;

[0028] Figure 3 It is a schematic diagram of the process of step S3.

[0029] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0030] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0031] In the description of the present invention, it is necessary to understand that terms such as “upper”, “lower”, “front”, “back”, “left”, “right”, “top”, “bottom”, “inside” and “outside” indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the referred system or element must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as a limitation on the present invention.

[0032] Example 1, see Figure 1 The present invention provides a smart agricultural pest monitoring method, which comprises the following steps:

[0033] Step S1: Data collection: Collect historical agricultural pest and disease monitoring data;

[0034] Step S2: Data preprocessing: performing data cleaning, data conversion, standardization and data optimization on the collected data to construct a time series data set;

[0035] Step S3: constructing a pest risk assessment model; by defining a dynamic adjustment function, designing a pest dynamic trend loss and introducing retrospective residual optimization, the pest risk assessment model is constructed;

[0036] Step S4: agricultural pest and disease monitoring: agricultural pest and disease monitoring is implemented based on the established pest and disease risk assessment model and the agricultural pest and disease monitoring data collected in real time.

[0037] Example 2, see Figure 1This embodiment is based on the above embodiment. In step S1, the historical agricultural pest and disease monitoring data includes pest and disease type data, time, geospatial data, climate and environmental data, crop growth data, pesticide data and pest and disease risk assessment results; the pest and disease risk assessment results include normal, mild danger, moderate danger and severe danger; the pest and disease risk assessment results are used as data labels; the pest and disease type data include pest and disease type, pest and disease occurrence time and pest and disease growth stage; the crop growth data include crop type, crop growth stage and crop health status; the climate and environmental data include temperature, humidity, precipitation, wind speed and wind direction; the geospatial data include farmland geographical location, field type and crop planting area type; the pesticide data include pesticide type, pesticide usage, pesticide usage time and application method.

[0038] Example 3, see Figure 1 , this embodiment is based on the above embodiment. In step S2, data cleaning is to process missing values ​​and duplicate values; data conversion is to convert the collected data into vector form; standardization processing is to standardize the data based on the maximum and minimum normalization method; data optimization is that the sensor will be affected by environmental changes, equipment errors and failures, resulting in the collected data containing non-Gaussian noise, so an environmental noise suppression metric is constructed; to reduce the impact of noise, especially to process abnormal data from sensors and the external environment; if the environmental noise suppression metric values ​​of the data and the other 90% of the data are lower than the metric threshold, the data is deleted; the environmental noise suppression metric value between the data It is expressed as: ; Where A and B are different sample data; is the normalization coefficient; N is the total number of samples; and They are shape parameter and scale parameter respectively; is the gamma function; is the time decay factor; is the sample sampling time; t is the current time; is the weight decay coefficient; d(·) is the Euclidean distance; is the current sample; is the average sample of the dataset.

[0039] By performing the above operations, the general agricultural pest and disease monitoring methods are affected by noise, environmental changes, equipment errors and failures, which leads to poor accuracy of agricultural pest and disease monitoring. This solution introduces environmental noise suppression metrics, identifies and removes abnormal monitoring data, introduces time attenuation factors and Euclidean distances between samples, so that noise suppression not only considers the similarity between samples, but also the dynamic changes in time and space, thereby improving the accuracy and real-time processing capabilities of agricultural pest and disease monitoring.

[0040] Example 4, see Figure 1 and Figure 3 Based on the above embodiment, in step S3, building a pest risk assessment model specifically includes the following steps:

[0041] Step S31: Preliminary construction; predict the pest risk assessment value based on LSTM, and optimize the pest risk assessment value by defining a dynamic adjustment function, and then build a pest risk assessment model; in order to more effectively calculate the pest risk assessment of pests and diseases, and avoid repeatedly calculating the inverse matrix of the gain matrix in a large-scale monitoring system, define a dynamic adjustment function , expressed as: ; ; The updated formula for pest and disease risk assessment after dynamic adjustment function optimization is: ;in, is the sample data; is the prediction error of LSTM for sample data; x is the predicted pest risk value of LSTM; and are the pest and disease risk assessment values ​​before and after optimization, respectively; is the adaptive weight function; is the noise threshold; is the uncertainty of the i-th sample; is the weight decay coefficient; is the scale factor;

[0042] Step S32: Design the dynamic trend loss of pests and diseases; if the model predicts that the pest risk in the region will increase in the future, and the actual situation is indeed so, the prediction is considered correct; if the model predicts that the risk will increase, but the risk actually decreases, a penalty is imposed; the dynamic trend loss ML of pests and diseases is expressed as: ; Where N is the total sample size; and are the actual pest risk value and the predicted actual pest risk value of the i-th sample respectively; and are the actual pest risk value and the predicted actual pest risk value of the i-1th sample respectively; is a smoothing term; is the trend mutation penalty factor; is the amount of risk change; Q is used to control the rate of exponential decay;

[0043] Step S33: Review residual optimization; in order to further improve the monitoring accuracy, the residual function is designed for nonlinear optimization, and the residual function is reviewed. It is expressed as: ;in, is the prediction error of the i-th sample; is the threshold used to control the penalty residual; and Used to adjust nonlinear shapes; is the residual constraint factor; m is the number of samples reviewed; j is the sample index; is the prediction error of the ijth sample;

[0044] Step S34: construct the final loss function; the loss function TL of the pest risk assessment model is expressed as: ;in, is the loss weight;

[0045] Step S35: model determination; the time series data set corresponding to the historical agricultural pest and disease monitoring data is divided into a test set and a training set in advance; the pest and disease risk assessment model uses a gradient descent algorithm to update parameters; when the pest and disease risk assessment model converges with the training set loss, the pest and disease risk assessment model training is completed; a prediction threshold is set in advance, and when the prediction accuracy of the trained pest and disease risk assessment model for the test set is higher than the prediction threshold, the pest and disease risk assessment model is established; otherwise, the data set is re-divided, the initial parameters are adjusted, and re-training is performed.

[0046] By performing the above operations, in view of the problem that general agricultural pest and disease monitoring methods ignore errors and uncertainties, and the risk assessment of future pests and diseases is inaccurate or too rough when facing high volatility and complex environments, which leads to poor pest and disease monitoring effects, this scheme optimizes the pest and disease risk assessment value by defining a dynamic adjustment function, dynamically adjusts the assessment value according to the prediction error of different sample data, and improves the prediction accuracy; by designing the dynamic trend loss of pests and diseases and introducing retrospective residual optimization, the error is continuously reduced in multiple predictions, thereby improving the agricultural pest and disease monitoring effect.

[0047] Example 5, see Figure 1 This embodiment is based on the above embodiment. In step S4, agricultural pest and disease monitoring is based on the established pest and disease risk assessment model, and agricultural pest and disease monitoring data is collected in real time. Agricultural pest and disease monitoring is realized based on the pest and disease risk assessment results predicted by the model; when the pest and disease risk assessment result predicted by the model is a medium risk or a high risk, early warning processing is performed.

[0048] Example 6, see Figure 2 , this embodiment is based on the above embodiment, and the present invention provides a smart agricultural pest monitoring system, including a data acquisition module, a data preprocessing module, a pest risk assessment model building module and an agricultural pest monitoring module;

[0049] The data acquisition module collects historical agricultural pest monitoring data and sends the data to the data preprocessing module;

[0050] The data preprocessing module performs data cleaning, data conversion, standardization processing and data optimization on the collected data; and sends the data to the pest risk assessment model construction module;

[0051] The pest risk assessment model building module defines a dynamic adjustment function, designs the pest dynamic trend loss and introduces retrospective residual optimization, thereby building a pest risk assessment model; and sends the data to the agricultural pest monitoring module;

[0052] The agricultural pest and disease monitoring module realizes agricultural pest and disease monitoring based on the agricultural pest and disease monitoring data collected in real time based on the established pest and disease risk assessment model.

[0053] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0054] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

[0055] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. A smart agricultural pest monitoring method, characterized by: The method comprises the following steps: Step S1: Data collection: Collect historical agricultural pest and disease monitoring data; Step S2: Data preprocessing: performing data cleaning, data conversion, standardization and data optimization on the collected data to construct a time series data set; Step S3: constructing a pest and disease risk assessment model; Step S4: agricultural pest monitoring: realizing agricultural pest monitoring based on the established pest risk assessment model and the agricultural pest monitoring data collected in real time; In step S2, the data optimization is to delete the data if the environmental noise suppression metric value of the data and the other 90% of the data is lower than the metric threshold; the environmental noise suppression metric value between the data is It is expressed as: ; Where A and B are different sample data; is the normalization coefficient; N is the total number of samples; and They are shape parameter and scale parameter respectively; is the gamma function; is the time decay factor; is the sample sampling time; t is the current time; is the weight decay coefficient; d(·) is the Euclidean distance; is the current sample; is the average sample of the dataset.

2. The method for monitoring pests and diseases in smart agriculture according to claim 1, characterized in that: In step S3, the construction of the pest risk assessment model specifically includes the following steps: Step S31: Preliminary construction: predict the pest risk assessment value based on LSTM, and optimize the pest risk assessment value by defining a dynamic adjustment function, thereby constructing a pest risk assessment model; define a dynamic adjustment function , expressed as: ; ; The updated formula for pest and disease risk assessment after dynamic adjustment function optimization is: ;in, is the sample data; is the prediction error of LSTM for sample data; x is the predicted pest risk value of LSTM; and are the pest and disease risk assessment values ​​before and after optimization, respectively; is the adaptive weight function; is the noise threshold; is the uncertainty of the i-th sample; is the weight decay coefficient; is the scale factor; Step S32: Design the dynamic trend loss of pests and diseases; Step S33: review residual optimization; Step S34: construct the final loss function; the loss function TL of the pest risk assessment model is expressed as: ;in, is the loss weight; ML is the dynamic trend loss of pests and diseases; is the retrospective residual function; Step S35: model determination; the time series data set corresponding to the historical agricultural pest and disease monitoring data is divided into a test set and a training set in advance; the pest and disease risk assessment model uses a gradient descent algorithm to update parameters; when the pest and disease risk assessment model converges with the training set loss, the pest and disease risk assessment model training is completed; a prediction threshold is set in advance, and when the prediction accuracy of the trained pest and disease risk assessment model for the test set is higher than the prediction threshold, the pest and disease risk assessment model is established; otherwise, the data set is re-divided, the initial parameters are adjusted, and re-training is performed.

3. A smart agricultural pest monitoring method according to claim 2, characterized in that: In step S32, the pest dynamic trend loss ML is expressed as: ; Where N is the total sample size; and are the actual pest risk value and the predicted actual pest risk value of the i-th sample respectively; and are the actual pest risk value and the predicted actual pest risk value of the i-1th sample respectively; is a smoothing term; is the trend mutation penalty factor; is the risk change.

4. The method for monitoring pests and diseases in smart agriculture according to claim 3, characterized in that: In step S33, the residual function is reviewed It is expressed as: ;in, is the prediction error of the i-th sample; is the threshold used to control the penalty residual; and Used to adjust nonlinear shapes; is the residual constraint factor; m is the number of samples reviewed; Q is used to control the rate of exponential decay; j is the sample index; is the prediction error of the ijth sample.

5. The method for monitoring pests and diseases in smart agriculture according to claim 4, characterized in that: In step S1, the historical agricultural pest and disease monitoring data includes pest and disease type data, time, geographic space data, climate and environment data, crop growth data, pesticide data and pest and disease risk assessment results; the pest and disease risk assessment results include normal, mild danger, moderate danger and severe danger; the pest and disease risk assessment results are used as data labels.

6. The method for monitoring pests and diseases in smart agriculture according to claim 1, characterized in that: In step S4, the agricultural pest and disease monitoring is based on the established pest and disease risk assessment model, and the agricultural pest and disease monitoring data is collected in real time, and the agricultural pest and disease monitoring is realized based on the pest and disease risk assessment results predicted by the model.

7. The method for monitoring pests and diseases in smart agriculture according to claim 1, characterized in that: In step S2, the data cleaning is to process missing values ​​and duplicate values; the data conversion is to convert the collected data into a vector form; and the standardization processing is to standardize the data based on the maximum and minimum normalization method.

8. A smart agricultural pest monitoring system, used to implement a smart agricultural pest monitoring method as claimed in any one of claims 1 to 7, characterized in that: It includes data collection module, data preprocessing module, pest and disease risk assessment model building module and agricultural pest and disease monitoring module; The data acquisition module collects historical agricultural pest monitoring data and sends the data to the data preprocessing module; The data preprocessing module performs data cleaning, data conversion, standardization and data optimization on the collected data to construct a time series data set; And send the data to the pest risk assessment model building module; The pest risk assessment model building module defines a dynamic adjustment function, designs the pest dynamic trend loss and introduces retrospective residual optimization, thereby building a pest risk assessment model; and sends the data to the agricultural pest monitoring module; The agricultural pest and disease monitoring module realizes agricultural pest and disease monitoring based on the agricultural pest and disease monitoring data collected in real time based on the established pest and disease risk assessment model.

Images