Intelligent management and analysis method and system for pesticide laboratory data

Abstract

The invention discloses a pesticide laboratory data intelligent management and analysis method and system. Pesticide data source features obtained through processing are obtained, pesticide data management features are obtained through processing in combination with pesticide detection equipment data obtained through processing and pesticide laboratory environment deviation features, and a corresponding pesticide data intelligent management analysis method is marked. And constructing a machine learning model based on pesticide data feature improvement by utilizing pesticide data management features and a pesticide data intelligent management analysis method. A cross-modal fusion learning model is constructed by comprehensively processing experimental fields, actual pesticide use conditions and related data of detection equipment, the level of a traditional classification and recognition model in pesticide laboratory data management and analysis is improved by constructing a machine learning model improved on the basis of pesticide data features, and a pesticide laboratory data management and analysis result is obtained by combining laboratory environment data. The intelligent management and analysis accuracy of the pesticide data is improved, so that the data management of the pesticide laboratory is more efficient, and the intelligence of the data management of the laboratory is improved.

Description

Technical Field

[0001] The system of this invention is used in the field of industrial information and data processing technology research and development, and in particular relates to a method and system for intelligent management and analysis of pesticide laboratory data. Background Technology

[0002] A pesticide laboratory data intelligent management and analysis system is a core tool for modern laboratory management. By integrating advanced information technology, artificial intelligence algorithms, and automated equipment, it achieves comprehensive, efficient, and intelligent management of pesticide testing data. Existing pesticide laboratories using LIMS systems possess the following core functions:

[0003] Sample lifecycle management: Tracking samples throughout their entire lifecycle, from receipt, allocation, preparation, testing to storage and destruction, to ensure complete and traceable sample information.

[0004] Automatic data acquisition and integration: Connects to various testing devices via interfaces to automatically acquire experimental data and perform unified format conversion.

[0005] Standardized process management: Establish standard operating procedures to ensure standardized operation of each testing step.

[0006] However, in existing technologies, there are still challenges in how to acquire and process pesticide data for management and analysis, how to utilize the characteristics of pesticide data sources and combine them with relevant feature data to obtain corresponding intelligent management and analysis methods for pesticide data, and how to construct machine learning models based on pesticide data features for application in pesticide data management. Furthermore, there are still significant room for improvement in classifiers and data processing to enhance the performance of traditional classification and recognition models in pesticide laboratory data management and analysis, combine laboratory environmental data to improve the accuracy of intelligent management and analysis of pesticide data, and ultimately make pesticide laboratory data management more efficient and intelligent. Summary of the Invention

[0007] To address the aforementioned technical problems, this invention proposes a method and system for intelligent management and analysis of pesticide laboratory data.

[0008] In a first aspect of the present invention, a method for intelligent management and analysis of pesticide laboratory data is provided, the method comprising:

[0009] S1. The pesticide data source characteristics are obtained and processed through the pesticide data storage client, and the processed pesticide testing equipment data and pesticide laboratory environment deviation characteristics are obtained and processed. The corresponding pesticide data intelligent management and analysis method is marked by personnel.

[0010] S2. After dimensionless processing based on the pesticide data source characteristics, the pesticide detection equipment data, and the pesticide laboratory environment deviation characteristics, pesticide data management characteristics are obtained.

[0011] S3. Based on the pesticide data management characteristics and the pesticide data intelligent management and analysis method, construct a machine learning model based on pesticide data characteristics improvement;

[0012] S4. Based on the improved machine learning model based on pesticide data features, output the real-time intelligent management and analysis method for pesticide data based on the currently acquired pesticide data.

[0013] Furthermore, the pesticide data source characteristics include the detection of pesticide data source, pesticide data source location characteristics, and pesticide data physicochemical characteristics.

[0014] Furthermore, the pesticide testing equipment data is calculated based on the usage frequency, maintenance frequency, and environmental deviation values ​​of each testing equipment.

[0015] Furthermore, the deviation characteristics of the pesticide laboratory environment are obtained by calculating and characterizing the pesticide collection temperature, transportation time, and laboratory ambient temperature at the time of pesticide collection at the pesticide location.

[0016] Furthermore, the pesticide data management features are obtained by performing dimensionless processing based on the pesticide data source characteristics, the pesticide testing equipment data, and the pesticide laboratory environment deviation characteristics, and the pesticide data management features are obtained by using a vector value connection method.

[0017] Furthermore, the machine learning model based on pesticide data features is improved by merging Fisher criterion classifiers trained on experimental field and non-experimental field datasets to obtain an intelligent management and analysis model for pesticide laboratory data.

[0018] It also provides a pesticide laboratory data intelligent management and analysis system, which includes a pesticide data source acquisition module, a pesticide testing equipment data processing module, a pesticide laboratory environmental deviation retrieval module, a pesticide laboratory data intelligent management and analysis model construction module, and a pesticide data intelligent management and analysis module.

[0019] The pesticide data source acquisition module: acquires and processes pesticide data source characteristics through the pesticide data storage client;

[0020] The pesticide testing equipment data processing module is used to acquire and process the pesticide testing equipment data.

[0021] The pesticide laboratory environment deviation retrieval module is used to acquire and process pesticide laboratory environment deviation characteristics.

[0022] The pesticide laboratory data intelligent management and analysis model construction module: Based on the pesticide data source characteristics, the pesticide testing equipment data, and the pesticide laboratory environment deviation characteristics, dimensionless processing is performed to obtain pesticide data management characteristics, and corresponding pesticide data intelligent management and analysis methods for personnel labels are obtained. Based on the pesticide data management characteristics and the pesticide data intelligent management and analysis methods, a machine learning model based on pesticide data characteristics improvement is constructed.

[0023] The pesticide data intelligent management and analysis module: based on the improved machine learning model based on pesticide data features, it outputs a real-time intelligent management and analysis method for the currently acquired pesticide data.

[0024] Furthermore, the pesticide testing equipment data is calculated based on the usage frequency, maintenance frequency, and environmental deviation values ​​of each testing equipment.

[0025] Furthermore, the machine learning model based on pesticide data features is improved by merging Fisher criterion classifiers trained on experimental field and non-experimental field datasets to obtain an intelligent management and analysis model for pesticide laboratory data.

[0026] Therefore, the beneficial effects of this invention are as follows: by acquiring and processing pesticide data source characteristics, and combining them with processed pesticide testing equipment data and pesticide laboratory environmental deviation characteristics, the calculation and acquisition of pesticide laboratory environmental deviation characteristics reduces resource utilization and pesticide data testing costs in the pesticide data management process. Processing yields pesticide data management characteristics, which are then used to label corresponding intelligent pesticide data management and analysis methods. A machine learning model based on improved pesticide data characteristics is constructed using these characteristics and methods. Multiple types of models are built using pesticide data from different sources, such as experimental field data and non-experimental field pesticide data. Furthermore, this invention constructs a cross-modal fusion learning model by comprehensively processing experimental field and actual pesticide usage data, as well as testing equipment-related data. By constructing a machine learning model based on improved pesticide data characteristics, the level of traditional classification and recognition models in pesticide laboratory data management and analysis is improved. Combined with laboratory environmental data, the accuracy of intelligent management and analysis of pesticide data is enhanced, making pesticide laboratory data management more efficient and thus improving the intelligence of laboratory data management.

[0027] Further embodiments and improvements of the present invention will be described in conjunction with the accompanying drawings and specific examples. Attached Figure Description

[0028] Figure 1 This is a flowchart of the intelligent management and analysis method for pesticide laboratory data of the present invention;

[0029] Figure 2 This is a schematic diagram of the intelligent management and analysis system for pesticide laboratory data of the present invention;

[0030] Figure 3 This is an example diagram of a pesticide testing device in one of the embodiments of the present invention;

[0031] Figure 4 This is a schematic diagram illustrating the multi-classifier fusion decision-making principle in an embodiment of the present invention;

[0032] Figure 5 This is a schematic diagram of an electronic device structure for implementing the method of the present invention in an embodiment of the present invention. Detailed Implementation

[0033] The invention will now be further described in conjunction with the accompanying drawings and specific embodiments. The machine learning model used in this invention is a machine learning model improved based on pesticide data features. The fusion machine learning model used for classification is also a personalized model for data processing related to pesticide laboratory data scenarios.

[0034] like Figure 2 As shown, this invention belongs to the field of industrial information and data processing technology research and development in agricultural laboratories, and belongs to information processing services such as industrial information and data processing, and therefore belongs to Internet and cloud computing, big data services.

[0035] In a first aspect of the present invention, a method for intelligent management and analysis of pesticide laboratory data is provided, the method comprising:

[0036] S1. The pesticide data source characteristics are obtained and processed through the pesticide data storage client, and the processed pesticide testing equipment data and pesticide laboratory environment deviation characteristics are obtained and processed. The corresponding pesticide data intelligent management and analysis method is marked by personnel.

[0037] S2. After dimensionless processing based on the pesticide data source characteristics, the pesticide detection equipment data, and the pesticide laboratory environment deviation characteristics, pesticide data management characteristics are obtained.

[0038] S3. Based on the pesticide data management characteristics and the pesticide data intelligent management and analysis method, construct a machine learning model based on pesticide data characteristics improvement;

[0039] S4. Based on the improved machine learning model based on pesticide data features, output the real-time intelligent management and analysis method for pesticide data based on the currently acquired pesticide data.

[0040] Furthermore, the pesticide data source characteristics include the detection of pesticide data source, pesticide data source location characteristics, and pesticide data physicochemical characteristics.

[0041] In managing and analyzing laboratory pesticide data, the inventors discovered that the data is closely related to its source, particularly the location of the pesticide being tested. For example, there are significant differences between experimental pesticides used in experimental fields and those used in actual farmland production. Here, excluding the possibility of different farmlands, the pesticide source characteristics obtained solely based on pesticide differences have a substantial impact on subsequent pesticide data management and analysis methods. To better characterize the influence of pesticide data source, this embodiment also considers the characteristics of the pesticide source location, namely the environmental conditions at the pesticide spraying site and the physicochemical characteristics of the sprayed pesticide. By comprehensively considering both the characteristics of the pesticide source location and the physicochemical characteristics of the pesticide itself, the pesticide data obtained in the laboratory can be better managed and analyzed.

[0042] In this embodiment, the pesticide data source is 1 if it comes from the experimental field, and 0.2 if it comes from outside the experimental field. This is an approximate feature value estimate made by the inventors based on observations of a large amount of data during pesticide data management in the laboratory. This is used for the subsequent classifier model to process the actual data. Further, in this embodiment, the pesticide data source characteristics include the average temperature, average humidity, and average precipitation over a period of time after pesticide spraying. It may also include the average soil pH, soil moisture, and soil humidity. To ensure the simplicity of the subsequent model data processing, this embodiment only uses the average temperature, average humidity, and average precipitation after a period of time following spraying. This time period can be set before the pesticide disappears, and is set to 3-5 days here. The physicochemical characteristics of the pesticide data are set based on relevant parameters before pesticide spraying. In this embodiment, density, pH value, and photolysis properties are used to characterize the physicochemical characteristics of the pesticide data.

[0043] At this point, the feature vector of the pesticide data source characteristics is represented as (detecting the pesticide data source, pesticide data source location characteristics, and pesticide data physicochemical characteristics). The pesticide data source detection, pesticide data source location characteristics, and pesticide data physicochemical characteristics are represented using the actual values ​​in the above embodiments. To facilitate the unified calculation of the subsequent model, the dimensions are removed.

[0044] Furthermore, the pesticide testing equipment data is calculated based on the usage frequency, maintenance frequency, and environmental deviation values ​​of each testing device:

[0045]

[0046] In the formula, D p Let i be the data of the pesticide testing equipment, n be the total number of pesticide testing equipment, and F be the data of the pesticide testing equipment. mi Let F be the maintenance frequency of the i-th pesticide testing equipment. ui Let T be the frequency of use of the i-th pesticide testing device.di Let T be the temperature during the detection of the i-th pesticide detection device. Ei The optimal temperature for detecting the i-th pesticide detection device.

[0047] The condition of pesticide testing equipment directly affects the accuracy, reliability, and traceability of pesticide data. Since pesticide data experiences significant wear and tear with each testing device during laboratory data management and analysis, data characterization of the testing lifespan of each device is used to comprehensively reduce the impact of equipment wear and tear on pesticide data, enabling precise management of pesticide data. Furthermore, environmental deviation values ​​are used to correct pesticide testing equipment data, resulting in more accurate management and analysis methods for pesticide data acquisition.

[0048] Furthermore, the deviation characteristics of the pesticide laboratory environment are obtained by calculating and characterizing the pesticide collection temperature, transportation time, and laboratory ambient temperature at the time of pesticide collection at the pesticide location:

[0049]

[0050] In the formula, F L Let be the deviation feature of the pesticide laboratory environment, and k be the feature correction coefficient to ensure that the deviation feature of the pesticide laboratory environment deviates by an order of magnitude from the feature of the pesticide data source, thereby reducing the generalization of the final classification result of the model. t represents the transportation time, after dimensionless processing. a T is the pesticide collection temperature during pesticide extraction. e This refers to the laboratory ambient temperature.

[0051] Current pesticide research often requires multiple protective environments to preserve pesticides in experimental or production fields for data acquisition and testing. This significantly wastes pesticide data testing resources. This invention utilizes the characteristics of pesticide laboratory environmental deviations to intelligently manage and analyze pesticide data acquisition, thereby reducing the utilization rate of environmental preservation equipment, lowering resource consumption, and achieving green pesticide data management and analysis. Pesticide testing is often affected by ambient temperature and the residence time from the outside environment to the laboratory. Therefore, this invention uses pesticide collection temperature, transportation time, and laboratory ambient temperature to characterize and calculate the characteristics of pesticide laboratory environmental deviations.

[0052] Furthermore, the pesticide data management features are obtained by performing dimensionless processing based on the pesticide data source characteristics, the pesticide testing equipment data, and the pesticide laboratory environment deviation characteristics, and the pesticide data management features are obtained by using a vector value connection method.

[0053] One of the characteristic vectors representing the pesticide data management features in this embodiment is (detecting pesticide data source, pesticide data source location characteristics, pesticide data physicochemical characteristics, pesticide testing equipment data, and pesticide laboratory environment deviation characteristics), the specific data of which are obtained by the method described above according to the present invention.

[0054] Furthermore, the machine learning model based on pesticide data features is improved by merging Fisher criterion classifiers trained on experimental field and non-experimental field datasets to obtain an intelligent management and analysis model for pesticide laboratory data.

[0055] Furthermore, the calculation formula for the intelligent management and analysis model of pesticide laboratory data is as follows:

[0056]

[0057] In the formula, A(X) represents the intelligent management and analysis method for pesticide data, and E... m M represents the amount of pesticide data in the experimental field dataset. t ω1 represents the total amount of pesticide data. T The normal vector of the Fisher criterion classifier trained on the experimental field dataset is perpendicular to the hyperplane, ω2. T X is the normal vector of the Fisher criterion classifier trained on the non-experimental field dataset, perpendicular to the hyperplane, and X is the pesticide data management feature of the input model.

[0058] In this embodiment, ω1 T To utilize the pesticide data management features corresponding to the pesticide data from the experimental field and the corresponding intelligent management and analysis method for pesticide data from the experimental field, the Fisher criterion classifier trained is used to obtain the normal vector perpendicular to the hyperplane, ω2. T The Fisher criterion classifier, trained using pesticide data management features corresponding to pesticide data from non-experimental fields and intelligent management analysis methods for pesticide data from non-experimental fields, is perpendicular to the normal vector of the hyperplane.

[0059] In this embodiment, since the accuracy of data features from different pesticide data sources varies significantly, and experimental fields are often insufficient for analyzing pesticide data, this invention trains different integrated intelligent management and analysis models for pesticide laboratory data by utilizing datasets of pesticide data from both experimental and non-experimental fields. This effectively eliminates the influence of different pesticide data sources, and threshold judgments are made using the classification feature values ​​obtained by the classifier to obtain the corresponding intelligent management and analysis method for pesticide data.

[0060] Furthermore, in this embodiment, the value corresponding to A(X) can be intelligently managed and analyzed for pesticide data. If the value corresponding to A(X) is less than 5, the corresponding pesticide data is considered to be real data and can be analyzed. If the value corresponding to A(X) is greater than or equal to 5, the corresponding pesticide data is considered to be fake data and cannot be analyzed. The threshold setting here is determined based on the judgment of the classifier output value and is not necessarily 5. It can also be set according to the needs of technicians, which will not be elaborated here.

[0061] The core value of fusing classifiers trained from different data sources lies in the complementary decision-making from multiple perspectives and levels. This classifier ensemble learning approach averages the error of individual models, reduces sensitivity to noise or outlier data, and mitigates the risk of overfitting. By learning across multiple data distributions, the model can capture more universal patterns, thus performing more stably when facing new environments or unknown data, and improving prediction accuracy and robustness.

[0062] It also provides a pesticide laboratory data intelligent management and analysis system, which includes a pesticide data source acquisition module, a pesticide testing equipment data processing module, a pesticide laboratory environmental deviation retrieval module, a pesticide laboratory data intelligent management and analysis model construction module, and a pesticide data intelligent management and analysis module.

[0063] The pesticide data source acquisition module: acquires and processes pesticide data source characteristics through the pesticide data storage client;

[0064] The pesticide testing equipment data processing module is used to acquire and process the pesticide testing equipment data.

[0065] The pesticide laboratory environment deviation retrieval module is used to acquire and process pesticide laboratory environment deviation characteristics.

[0066] The pesticide laboratory data intelligent management and analysis model construction module: Based on the pesticide data source characteristics, the pesticide testing equipment data, and the pesticide laboratory environment deviation characteristics, dimensionless processing is performed to obtain pesticide data management characteristics, and corresponding pesticide data intelligent management and analysis methods for personnel labels are obtained. Based on the pesticide data management characteristics and the pesticide data intelligent management and analysis methods, a machine learning model based on pesticide data characteristics improvement is constructed.

[0067] The pesticide data intelligent management and analysis module: based on the improved machine learning model based on pesticide data features, it outputs a real-time intelligent management and analysis method for the currently acquired pesticide data.

[0068] Furthermore, the pesticide testing equipment data is calculated based on the usage frequency, maintenance frequency, and environmental deviation values ​​of each testing device:

[0069]

[0070] In the formula, D p Let i be the data of the pesticide testing equipment, n be the total number of pesticide testing equipment, and F be the data of the pesticide testing equipment. mi Let F be the maintenance frequency of the i-th pesticide testing equipment. ui Let T be the frequency of use of the i-th pesticide testing device. di Let T be the temperature during the detection of the i-th pesticide detection device. Ei The optimal temperature for detecting the i-th pesticide detection device.

[0071] The condition of pesticide testing equipment directly affects the accuracy, reliability, and traceability of pesticide data. Since pesticide data experiences significant wear and tear with each testing device during laboratory data management and analysis, data characterization of the testing lifespan of each device is used to comprehensively reduce the impact of equipment wear and tear on pesticide data, enabling precise management of pesticide data. Furthermore, environmental deviation values ​​are used to correct pesticide testing equipment data, resulting in more accurate management and analysis methods for pesticide data acquisition.

[0072] Furthermore, the machine learning model based on pesticide data features is improved by merging Fisher criterion classifiers trained on experimental field and non-experimental field datasets to obtain an intelligent management and analysis model for pesticide laboratory data.

[0073] The calculation formula for the intelligent management and analysis model of pesticide laboratory data is as follows:

[0074]

[0075] In the formula, A(X) represents the intelligent management and analysis method for pesticide data, and E... m M represents the amount of pesticide data in the experimental field dataset. t ω1 represents the total amount of pesticide data. T The normal vector of the Fisher criterion classifier trained on the experimental field dataset is perpendicular to the hyperplane, ω2. T X is the normal vector of the Fisher criterion classifier trained on the non-experimental field dataset, perpendicular to the hyperplane, and X is the pesticide data management feature of the input model.

[0076] In this embodiment, ω1 T To utilize the pesticide data management features corresponding to the pesticide data from the experimental field and the corresponding intelligent management and analysis method for pesticide data from the experimental field, the Fisher criterion classifier trained is used to obtain the normal vector perpendicular to the hyperplane, ω2. TThe Fisher criterion classifier, trained using pesticide data management features corresponding to pesticide data from non-experimental fields and intelligent management analysis methods for pesticide data from non-experimental fields, is perpendicular to the normal vector of the hyperplane.

[0077] In this embodiment, since the accuracy of data features from different pesticide data sources varies significantly, and experimental fields are often insufficient for analyzing pesticide data, this invention trains different integrated intelligent management and analysis models for pesticide laboratory data by utilizing datasets of pesticide data from both experimental and non-experimental fields. This effectively eliminates the influence of different pesticide data sources, and threshold judgments are made using the classification feature values ​​obtained by the classifier to obtain the corresponding intelligent management and analysis method for pesticide data.

[0078] Furthermore, in this embodiment, the value corresponding to A(X) can be intelligently managed and analyzed for pesticide data. If the value corresponding to A(X) is less than 5, the corresponding pesticide data is considered to be real data and can be analyzed. If the value corresponding to A(X) is greater than or equal to 5, the corresponding pesticide data is considered to be fake data and cannot be analyzed. The threshold setting here is determined based on the judgment of the classifier output value and is not necessarily 5. It can also be set according to the needs of technicians, which will not be elaborated here.

[0079] The core value of fusing classifiers trained from different data sources lies in the complementary decision-making from multiple perspectives and levels. This classifier ensemble learning approach averages the error of individual models, reduces sensitivity to noise or outlier data, and mitigates the risk of overfitting. By learning across multiple data distributions, the model can capture more universal patterns, thus performing more stably when facing new environments or unknown data, and improving prediction accuracy and robustness.

[0080] Therefore, the beneficial effects of this invention are as follows: by acquiring and processing pesticide data source characteristics, and combining them with processed pesticide testing equipment data and pesticide laboratory environmental deviation characteristics, the calculation and acquisition of pesticide laboratory environmental deviation characteristics reduces resource utilization and pesticide data testing costs in the pesticide data management process. Processing yields pesticide data management characteristics, which are then used to label corresponding intelligent pesticide data management and analysis methods. A machine learning model based on improved pesticide data characteristics is constructed using these characteristics and methods. Multiple types of models are built using pesticide data from different sources, such as experimental field data and non-experimental field pesticide data. Furthermore, this invention constructs a cross-modal fusion learning model by comprehensively processing experimental field and actual pesticide usage data, as well as testing equipment-related data. By constructing a machine learning model based on improved pesticide data characteristics, the level of traditional classification and recognition models in pesticide laboratory data management and analysis is improved. Combined with laboratory environmental data, the accuracy of intelligent management and analysis of pesticide data is enhanced, making pesticide laboratory data management more efficient and thus improving the intelligence of laboratory data management.

[0081] Of course, it is understood that each embodiment of the present invention can achieve one of the effects individually, and the combination of multiple embodiments of the present invention can achieve all the above effects. However, it is not required that each embodiment of the present invention achieve all the above advantages and effects, because each embodiment of the present invention can constitute a separate technical solution and make one or more contributions to the prior art. Each embodiment of the present invention is not affected, and the solution can still be implemented even if one embodiment is deleted.

[0082] For any module structures not specifically defined in this invention, the existing technical descriptions shall prevail. The prior art mentioned in the foregoing background and specific embodiments sections can be considered part of this invention and used to understand the meaning of certain technical features or parameters.

Claims

1. A method for intelligent management and analysis of pesticide laboratory data, characterized in that, The method includes: S1. The pesticide data source characteristics are obtained and processed through the pesticide data storage client, and the processed pesticide testing equipment data and pesticide laboratory environment deviation characteristics are obtained and processed. The corresponding pesticide data intelligent management and analysis method is marked by personnel. S2. After dimensionless processing based on the pesticide data source characteristics, the pesticide detection equipment data, and the pesticide laboratory environment deviation characteristics, pesticide data management characteristics are obtained. S3. Based on the pesticide data management characteristics and the pesticide data intelligent management and analysis method, construct a machine learning model based on pesticide data characteristics improvement; S4. Based on the improved machine learning model based on pesticide data features, output the real-time intelligent management and analysis method for pesticide data based on the currently acquired pesticide data.

2. The intelligent management and analysis method for pesticide laboratory data as described in claim 1, characterized in that: The characteristics of pesticide data sources include the source of the pesticide data, the location of the pesticide data source, and the physicochemical characteristics of the pesticide data.

3. The intelligent management and analysis method for pesticide laboratory data as described in claim 2, characterized in that: The pesticide testing equipment data is calculated based on the usage frequency, maintenance frequency, and environmental deviation values ​​of each testing equipment.

4. The intelligent management and analysis method for pesticide laboratory data as described in claim 3, characterized in that: The deviation characteristics of the pesticide laboratory environment are obtained by calculating and characterizing the pesticide collection temperature, transportation time, and laboratory ambient temperature at the time of pesticide collection at the pesticide location.

5. The intelligent management and analysis method for pesticide laboratory data as described in claim 1 or 3, characterized in that: The pesticide data management features are obtained by performing dimensionless processing based on the pesticide data source features, the pesticide testing equipment data, and the pesticide laboratory environment deviation features, and the pesticide data management features are obtained by using a vector value connection method.

6. The intelligent management and analysis method for pesticide laboratory data as described in claim 1 or 4, characterized in that: The machine learning model based on pesticide data features is obtained by merging Fisher criterion classifiers trained on experimental field and non-experimental field datasets to form an intelligent management and analysis model for pesticide laboratory data.

7. A pesticide laboratory data intelligent management and analysis system, comprising a pesticide data source acquisition module, a pesticide testing equipment data processing module, a pesticide laboratory environmental deviation retrieval module, a pesticide laboratory data intelligent management and analysis model construction module, and a pesticide data intelligent management and analysis module, characterized in that: The pesticide data source acquisition module: acquires and processes pesticide data source characteristics through the pesticide data storage client; The pesticide testing equipment data processing module is used to acquire and process the pesticide testing equipment data. The pesticide laboratory environment deviation retrieval module is used to acquire and process pesticide laboratory environment deviation characteristics. The pesticide laboratory data intelligent management and analysis model construction module: Based on the pesticide data source characteristics, the pesticide testing equipment data, and the pesticide laboratory environment deviation characteristics, dimensionless processing is performed to obtain pesticide data management characteristics, and corresponding pesticide data intelligent management and analysis methods for personnel labels are obtained. Based on the pesticide data management characteristics and the pesticide data intelligent management and analysis methods, a machine learning model based on pesticide data characteristics improvement is constructed. The pesticide data intelligent management and analysis module: based on the improved machine learning model based on pesticide data features, it outputs a real-time intelligent management and analysis method for the currently acquired pesticide data.

8. The intelligent management and analysis system for pesticide laboratory data as described in claim 7, characterized in that: The pesticide testing equipment data is calculated based on the usage frequency, maintenance frequency, and environmental deviation values ​​of each testing equipment.

9. The intelligent management and analysis system for pesticide laboratory data as described in claim 7, characterized in that: The machine learning model based on pesticide data features is obtained by merging Fisher criterion classifiers trained on experimental field and non-experimental field datasets to form an intelligent management and analysis model for pesticide laboratory data.

Images