Method and system for pre-estimating and optimizing dispersion stability of herbicide suspending agent

Abstract

The invention belongs to the technical field of agriculture, and discloses a method and a system for estimating and optimizing dispersion stability of a herbicide suspending agent, which are used for estimating dispersion stability by simulating various environmental conditions to accelerate the degradation process of the suspending agent and measuring turbidity data in real time, and can be used for quickly and accurately estimating the dispersion stability.

Description

Technical Field

[0001] The present application relates to the field of agricultural technologies, and more particularly, to a method and system for predicting and optimizing the dispersion stability of a herbicide suspension concentrate. Background Art

[0002] Driven by the dual imperatives of environmental protection and sustainable agricultural development, environmentally friendly herbicide suspension concentrates, such as biogenic herbicides, plant essential oil-based herbicides, natural organic acid-based herbicides, amino acid-based herbicides, and polysaccharide-based herbicides, have gradually become an important part of modern agricultural production systems. In the face of the market's urgent need for environmentally friendly pesticide products, many small pesticide enterprises have actively invested in research and development resources and are committed to developing new herbicide products mainly composed of natural or biodegradable components. Such environmentally friendly suspension concentrates pay more attention to environmental friendliness in formulation design and strive to minimize the potential adverse impacts on the ecological environment.

[0003] However, compared with traditional herbicides, the problem of dispersion stability of environmentally friendly herbicide suspension concentrates is particularly prominent. Dispersion stability is not only directly related to the full release of the active ingredients of herbicides and is a key factor in maintaining the product shelf life and ensuring product quality, but also directly affects the reliability and persistence of field weed control effects. For small pesticide enterprises, ensuring the quality of each batch of products leaving the factory is of utmost importance, and strict quality control before listing is an indispensable link. Therefore, how to quickly and accurately evaluate the dispersion stability of environmentally friendly herbicide suspension concentrates under various complex environmental conditions has become a key technical problem in product quality control. Although the traditional long-term storage experiment is recognized as the gold standard method for evaluating stability, its inherent defect of being time-consuming severely restricts the ability of small enterprises to quickly adjust product formulations and flexibly respond to market changes. Especially in the face of complex and variable field environmental factors, such as high temperature, high humidity, and irrigation water of different hardness, traditional methods often fail to provide effective stability feedback information in a short time. The current industry urgently needs a simple, fast, and reliable dispersion stability evaluation method that can efficiently simulate various field extreme environments under laboratory conditions, so as to provide timely and effective technical support for the formulation optimization and quality control of environmentally friendly herbicide suspension concentrates, thereby effectively ensuring the healthy development of green agriculture and the stable reliability of field weed control effects. However, there is still a lack of a dispersion stability evaluation technology that can simultaneously take into account rapidity, accuracy, and environmental adaptability in the current market, and small pesticide enterprises still face double challenges of efficiency and cost in the quality control of environmentally friendly herbicide products.

[0004] In view of the above problems, there is an urgent need for improvement in the existing technologies. Summary of the Invention

[0005] The purpose of this application is to provide a method and system for predicting and optimizing the dispersion stability of herbicide suspension concentrates, which can quickly and accurately predict the dispersion stability.

[0006] In the first aspect, this application provides a method for predicting and optimizing the dispersion stability of herbicide suspension concentrates, and the method includes: A1. Dilute the suspension concentrate to be tested with simulated irrigation water of multiple different hardness levels to form multiple initial suspension concentrate samples, and divide each initial suspension concentrate sample into multiple parallel samples to form multiple groups of parallel samples; each group of parallel samples contains parallel samples corresponding to simulated irrigation water of all hardness levels; A2. Place each group of parallel samples under different environmental stress conditions to accelerate the deterioration of the dispersion stability of the suspension concentrate, and during the deterioration process, measure the turbidity data of each parallel sample in real time; the environmental stress conditions include temperature conditions and humidity conditions; A3. Calculate the turbidity change rate of each parallel sample based on the measured turbidity data to evaluate the dispersion stability performance of each parallel sample; A4. Evaluate the comprehensive dispersion stability performance of the suspension concentrate to be tested based on the dispersion stability performance of each parallel sample; A5. Analyze the influence level of each environmental factor on the dispersion stability of the suspension concentrate to be tested based on the dispersion stability performance of each parallel sample and the corresponding simulated irrigation water hardness and environmental stress conditions; the environmental factors include irrigation water hardness, temperature, and humidity; A6. Output an evaluation report including the dispersion stability performance of each parallel sample, the comprehensive dispersion stability performance of the suspension concentrate to be tested, and the influence level of each environmental factor on the dispersion stability of the suspension concentrate to be tested.

[0007] Furthermore, this application also proposes that step A1 includes: A101. Configure simulated irrigation water of multiple different hardness levels according to the common water quality range of the expected use area; A102. Dilute the suspension concentrate to be tested to the target concentration respectively according to the configured simulated irrigation water of different hardness levels to obtain multiple initial suspension concentrate samples; A103. Measure each initial suspension concentrate sample and dispense it into multiple sample containers, with each sample container corresponding to a parallel sample, to form multiple groups of parallel samples; each group of parallel samples contains parallel samples corresponding to simulated irrigation water of all hardness levels.

[0008] Furthermore, this application also proposes that step A2 includes: A201. Obtain the historical meteorological data of the expected use area, and the historical meteorological data includes temperature data and humidity data; A202. Use a clustering algorithm to extract multiple typical environmental stress combinations from historical meteorological data; each typical environmental stress combination includes a temperature data and a humidity data; A203. Set the temperature and humidity of each channel of the environmental stress acceleration device so that the environmental stress conditions of each channel correspond to a typical environmental stress combination; A204. Place each parallel sample group in each channel of the environmental stress acceleration device respectively to accelerate the deterioration of the dispersion stability of the suspending agent; A205. During the deterioration process, measure the turbidity data of each parallel sample in real time.

[0009] Furthermore, the present application also proposes that step A202 includes: B1. Conduct clustering analysis on historical meteorological data using multiple clustering algorithms respectively to obtain multiple clustering results; the clustering algorithms include using the K-means clustering algorithm, hierarchical clustering algorithm, and DBSCAN clustering algorithm; B2. Evaluate multiple clustering results according to multiple evaluation indicators, and select the optimal clustering result; the evaluation indicators include silhouette coefficient, Calinski-Harabasz index, and Davies-Bouldin index; B3. Calculate the temperature mean and humidity mean in each cluster of the optimal clustering result to obtain the typical environmental stress combination of the corresponding cluster.

[0010] Furthermore, the present application also proposes that step B1 includes: B101. Conduct standardization processing on the temperature data and humidity data for historical meteorological data to eliminate the dimension difference; B102. Use the elbow method to determine the optimal number of clusters for the K-means clustering algorithm, and based on the determined optimal number of clusters, use the K-means clustering algorithm to conduct clustering analysis on the standardized historical meteorological data to generate multiple clusters and obtain the K-means algorithm clustering result; B103. For the hierarchical clustering algorithm, use the Wardlinkage method to calculate the inter-cluster distance, construct a hierarchical clustering dendrogram, and according to the set threshold or number of clusters, divide the dendrogram into multiple clusters to obtain the hierarchical clustering algorithm clustering result; B104. For the DBSCAN clustering algorithm, use the grid search method to determine the neighborhood radius and minimum sample number parameters, and based on the determined neighborhood radius and minimum sample number parameters, use the DBSCAN clustering algorithm to conduct clustering analysis on historical meteorological data to generate multiple clusters and obtain the DBSCAN algorithm clustering result.

[0011] Furthermore, the present application also proposes that step B2 includes: B201. Calculate the silhouette coefficient, Calinski-Harabasz index, and Davies-Bouldin index for multiple clustering results respectively; B202. For each evaluation index, perform normalization and reciprocal processing to map the values of each evaluation index to the interval [0, 1] to obtain the normalized evaluation index; B203. For each clustering result, perform weighted summation of the normalized silhouette coefficient, normalized Calinski-Harabasz index, and normalized Davies-Bouldin index to obtain a comprehensive score; B204. Select the clustering result with the highest comprehensive score as the optimal clustering result.

[0012] Furthermore, the present application also proposes that step A3 includes: A301. Extract the turbidity data of each parallel sample at different time points to form a turbidity time series; A302. For each turbidity time series, use the sliding window method to calculate the turbidity change rate; A303. For each parallel sample, calculate the average turbidity change rate of the entire deterioration process according to the calculated results of the corresponding turbidity change rate, and convert it into a dispersion stability performance evaluation index.

[0013] Furthermore, the present application also proposes that step A303 includes: For each parallel sample, extract the turbidity change rate time series and divide it into multiple time windows according to the time sequence; For each time window, calculate the average weighted value of the turbidity change rate; the farther the time window is from the initial moment of the deterioration process, the smaller the corresponding weight coefficient; For each parallel sample, accumulate the average weighted values of the turbidity change rate of each time window as the dispersion stability performance evaluation index.

[0014] Furthermore, the present application also proposes that step A4 includes: A401. Obtain the historical meteorological data of the expected use area and calculate the occurrence probability of each environmental stress condition in the expected use area according to the historical meteorological data; A402. Convert the occurrence probability of each environmental stress condition in the expected use area into a weight coefficient; A403. Calculate the weighted average dispersion stability performance according to the dispersion stability performance of each parallel sample and the corresponding weight coefficient as the comprehensive dispersion stability performance of the suspension agent to be tested.

[0015] In the second aspect, the present application also proposes a herbicide suspension agent dispersion stability prediction and optimization system, which includes: An environmental stress acceleration device has multiple channels, and the environmental stress conditions of each channel can be independently adjusted. Each channel is used to accelerate the deterioration of the suspension dispersion stability of different parallel sample groups under different environmental stress conditions. The environmental stress conditions include temperature conditions and humidity conditions. The parallel sample groups are obtained in the following way: Dilute the suspension to be tested with simulated irrigation water of multiple different hardnesses to form multiple initial suspension samples, and divide each initial suspension sample into multiple parallel samples to form multiple parallel sample groups. Each parallel sample group contains parallel samples corresponding to simulated irrigation water of all hardnesses; A turbidity detection device is used to measure the turbidity data of each parallel sample in real time during the deterioration process; A host computer is used to perform the following steps: According to the measured turbidity data, calculate the turbidity change rate of each parallel sample to evaluate the dispersion stability performance of each parallel sample; According to the dispersion stability performance of each parallel sample, evaluate the comprehensive dispersion stability performance of the suspension to be tested; According to the dispersion stability performance of each parallel sample, the hardness of the simulated irrigation water corresponding thereto, and the environmental stress conditions, analyze the influence level of each environmental factor on the dispersion stability of the suspension to be tested. The environmental factors include the hardness of the irrigation water, temperature, and humidity; Output an evaluation report including the dispersion stability performance of each parallel sample, the comprehensive dispersion stability performance, and the influence level of each environmental factor on the dispersion stability of the suspension to be tested.

[0016] Beneficial effects: A method and system for predicting and optimizing the dispersion stability of a herbicide suspension provided by this application can accelerate the deterioration process of the suspension by simulating various environmental conditions and measure the turbidity data in real time to evaluate the dispersion stability, and can quickly and accurately predict the dispersion stability. Description of the Drawings

[0017] Figure 1 It is a flowchart of the method for predicting and optimizing the dispersion stability of a herbicide suspension provided by an embodiment of this application.

[0018] Figure 2 It is a structural schematic diagram of the system for predicting and optimizing the dispersion stability of a herbicide suspension provided by an embodiment of this application.

[0019] Label description: 1. Environmental stress acceleration device; 2. Turbidity detection device; 3. Host computer. Detailed Embodiment

[0020] The technical solutions in this application will be clearly and completely described below with reference to the accompanying drawings in this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments. The components of this application usually described and illustrated in the accompanying drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of this application claimed, but merely represents the selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative efforts fall within the scope of protection of this application.

[0021] It should be noted that: similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of this application, terms such as "first" and "second" are only used for distinguishing descriptions and cannot be understood as indicating or implying relative importance.

[0022] Referring to Figure 1 , this application proposes an optimization method for predicting the dispersion stability of a herbicide suspension concentrate, and the method includes: A1. Dilute the suspension concentrate to be tested with simulated irrigation water of multiple different hardness levels to form multiple initial suspension concentrate samples, and divide each initial suspension concentrate sample into multiple parallel samples to form multiple groups of parallel samples; each group of parallel samples contains parallel samples corresponding to simulated irrigation water of all hardness levels; A2. Place each group of parallel samples under different environmental stress conditions to accelerate the deterioration of the dispersion stability of the suspension concentrate, and during the deterioration process, measure the turbidity data of each parallel sample in real time; the environmental stress conditions include temperature conditions and humidity conditions; A3. Calculate the turbidity change rate of each parallel sample based on the measured turbidity data to evaluate the dispersion stability performance of each parallel sample; A4. Evaluate the comprehensive dispersion stability performance of the suspension concentrate to be tested based on the dispersion stability performance of each parallel sample; A5. Analyze the influence levels of each environmental factor on the dispersion stability of the suspension concentrate to be tested based on the dispersion stability performance of each parallel sample and the corresponding simulated irrigation water hardness and environmental stress conditions; the environmental factors include irrigation water hardness, temperature, and humidity; A6. Output an evaluation report including the dispersion stability performance of each parallel sample, the comprehensive dispersion stability performance of the suspension concentrate to be tested, and the influence levels of each environmental factor on the dispersion stability of the suspension concentrate to be tested.

[0023] Among them, in step A1, the suspension agent to be tested is first diluted by a series of configured simulated irrigation waters, and the hardness gradient of the simulated irrigation waters covers the common water quality ranges of the expected use areas. For example, the hardness of the simulated irrigation waters can be set to three or more levels, namely low hardness, medium hardness, and high hardness. Each initial suspension agent sample is then divided into multiple equal parts and placed in different sample containers to form parallel samples, aiming to ensure the reliability of the experimental results. The parallel samples are grouped according to the hardness of the simulated irrigation waters, constituting multiple parallel sample groups, so that each parallel sample group contains parallel samples under all hardness conditions, facilitating subsequent comparative experiments under the same environmental stress.

[0024] Among them, in step A2, each parallel sample group is placed in different channels of the environmental stress acceleration device. For example, the environmental stress acceleration device has multiple mutually separated chambers, and the environmental stress conditions in each chamber are independently controllable. Then each chamber is a channel. The environmental stress conditions, namely temperature and humidity, are set to different combinations in different channels to simulate the meteorological conditions that may occur in the expected use areas. The environmental stress acceleration device can precisely control the temperature and humidity of each channel to achieve the purpose of accelerating the deterioration of the suspension agent samples. During the deterioration process of the suspension agent, the turbidity detection device measures the turbidity data of each parallel sample in real time, and the turbidity data reflects the change in the dispersion state of the suspension agent. The turbidity detection device can be a non-contact optical sensor to avoid interfering with the samples.

[0025] Among them, in step A3, the turbidity change rate is calculated by analyzing the real-time measured turbidity data. For example, the sliding window method can be used to calculate the change amount of the turbidity value per unit time to characterize the dispersion stability performance of the suspension agent. The smaller the turbidity change rate, the better the dispersion stability of the suspension agent.

[0026] Among them, in step A4, the comprehensive dispersion stability performance is evaluated based on the dispersion stability performance of each parallel sample. For example, according to the historical meteorological data of the expected use area, the probability of occurrence of various environmental stress conditions can be determined and used as a weight coefficient to perform weighted averaging on the dispersion stability performance under each environmental stress condition to obtain the comprehensive dispersion stability performance.

[0027] Among them, in step A5, the influence level of environmental factors on the dispersion stability is obtained by comparing the dispersion stability performance analysis under different environmental conditions. For example, statistical analysis methods such as analysis of variance (ANOVA), regression analysis, or correlation analysis can be used to evaluate the influence degree of each environmental factor on the dispersion stability performance. Then, according to the statistical analysis results, the influence level of each environmental factor on the dispersion stability of the suspending agent is determined. Specifically, statistical analysis software such as SPSS, SAS, or R language can be used to complete the data analysis to obtain the quantitative influence degree analysis results, and then compare the quantitative influence degree analysis results with the preset threshold to determine the influence level.

[0028] Among them, in step A6, the evaluation report is finally output. The report includes the evaluation results of the dispersion stability performance of each parallel sample, the comprehensive dispersion stability performance evaluation results of the suspending agent to be tested, and the analysis results of the influence level of each environmental factor on the dispersion stability, providing data support for the formulation optimization of the herbicide suspending agent. For example, a preset evaluation report template can be retrieved from the database, and the evaluation results of the dispersion stability performance of each parallel sample, the comprehensive dispersion stability performance evaluation results of the suspending agent to be tested, and the analysis results of the influence level of each environmental factor on the dispersion stability are added to the corresponding positions of the evaluation report template to obtain the final evaluation report.

[0029] Specifically, the working principle of this method for predicting and optimizing the dispersion stability of the herbicide suspending agent is as follows: by simulating various environmental factors that may be encountered in actual applications, the dispersion stability deterioration process of the herbicide suspending agent is accelerated, and the turbidity change is monitored in real time during the deterioration process to quickly evaluate the dispersion stability of the suspending agent under different environments. First, through step A1, suspending agent samples diluted in irrigation water with different hardness are prepared, simulating the differences in water quality in actual use. Subsequently, in step A2, these samples are placed under different temperature and humidity combinations to accelerate the deterioration of the suspending agent, and the deterioration process is dynamically monitored through real-time turbidity measurement. Step A3 quantifies the dispersion stability performance of the suspending agent by calculating the turbidity change rate. Step A4 comprehensively considers the dispersion stability under different environmental conditions to obtain the comprehensive dispersion stability performance of the suspending agent to be tested, more comprehensively reflecting the stability performance of the product in actual applications. Step A5 further analyzes the influence degree of irrigation water hardness, temperature, and humidity on the dispersion stability, providing a clear direction for formulation optimization. Finally, step A6 outputs an evaluation report, providing a scientific basis for the quality control and formulation improvement of the herbicide suspending agent. This method avoids the disadvantage of the long time-consuming traditional long-term storage experiment and realizes the rapid prediction and optimization of the dispersion stability of the herbicide suspending agent.

[0030] In some embodiments, step A1 includes: A101. Configure simulated irrigation water with multiple different hardness levels according to the common water quality range of the expected usage area. A102. Dilute the suspension agent to be tested to the target concentration respectively according to the configured simulated irrigation water with different hardness levels to obtain multiple initial suspension agent samples. A103. Measure each initial suspension agent sample and dispense it into multiple sample containers. Each sample container corresponds to a parallel sample, forming multiple groups of parallel samples. Each group of parallel samples includes parallel samples corresponding to simulated irrigation water with all hardness levels.

[0031] Among them, in step A101, the common water quality range of the expected usage area can be understood as the statistical distribution of the hardness of irrigation water in this area. For example, by collecting water quality monitoring data in the expected usage area for many years and analyzing to obtain the range of water hardness. Configuring simulated irrigation water with multiple different hardness levels can be specifically implemented as follows. For example, if the water hardness range in the expected usage area is mainly distributed between 50 ppm and 300 ppm, six kinds of simulated irrigation water with hardness levels of 50 ppm, 100 ppm, 150 ppm, 200 ppm, 250 ppm, and 300 ppm can be configured. The preparation method of the simulated irrigation water can refer to the method in "Quality Control Technology for Pesticide Suspension Concentrates" and use reagents such as calcium chloride, magnesium sulfate, and sodium bicarbonate for preparation.

[0032] Among them, in step A102, the target concentration refers to the dilution concentration of the herbicide suspension concentrate during actual field use. This concentration is usually recommended by pesticide manufacturers according to product characteristics and usage scenarios. Diluting the suspension agent to be tested to the target concentration aims to simulate the state of the herbicide during actual use.

[0033] Among them, in step A103, dispensing into multiple sample containers is to set up parallel samples. Multiple parallel samples are prepared under each hardness condition for subsequent dispersion stability tests under different environmental stress conditions to ensure the reliability of the experimental results. The construction method of the group of parallel samples is to combine one parallel sample under all hardness conditions together to form a group of parallel samples for convenient testing and comparison under the same environmental stress condition. Among them, the number of parallel samples obtained by dispensing each initial suspension agent sample (i.e., the number of groups of parallel samples) can be set according to actual needs. For example, it can be set according to the number of channels of the environmental stress acceleration device, so that all groups of parallel samples can be accelerated for deterioration simultaneously, saving time. In fact, it can also exceed the number of channels of the environmental stress acceleration device, but the accelerated deterioration of each group of parallel samples needs to be carried out in batches.

[0034] Specifically, through step A101, the hardness setting of the simulated irrigation water can truly reflect the water quality characteristics of the expected use area. Thus, the subsequent evaluation of dispersion stability can be more in line with the actual application scenario. Step A102 ensures that the suspension agent to be tested is diluted to the actual use concentration in the simulated irrigation water, further improving the authenticity of the experiment. Step A103 ensures the accuracy and repeatability of the experimental results by preparing parallel sample groups. Through the above steps, parallel sample groups can be prepared that can represent the water quality characteristics of the expected use area and simulate the actual use concentration, laying a foundation for the subsequent evaluation of the dispersion stability of the herbicide suspension agent and making the evaluation results more practically instructive.

[0035] In some embodiments, step A2 includes: A201. Obtain historical meteorological data of the expected use area, where the historical meteorological data includes temperature data and humidity data; A202. Use a clustering algorithm to extract multiple typical environmental stress combinations from the historical meteorological data; each typical environmental stress combination includes a temperature data and a humidity data; A203. Set the temperature and humidity of each channel of the environmental stress acceleration device so that the environmental stress conditions of each channel correspond to a typical environmental stress combination; A204. Place each parallel sample group in each channel of the environmental stress acceleration device respectively to accelerate the deterioration of the dispersion stability of the suspension agent; A205. During the deterioration process, measure the turbidity data of each parallel sample in real time.

[0036] Among them, in step A201, the historical meteorological data can be obtained from the public database of the meteorological bureau, and the time span of the data can be set to data of recent years or recent decades to ensure the representativeness of the data.

[0037] Among them, in step A202, a clustering algorithm is used to analyze historical meteorological data to extract representative environmental stress combinations. Specifically, historical meteorological data is first collected, and the data contains temperature and humidity information. Subsequently, various clustering algorithms, such as the K-means algorithm, hierarchical clustering algorithm, or DBSCAN algorithm, can be applied to identify patterns in the meteorological data and group similar meteorological conditions into several clusters. Each cluster represents a typical environmental stress state. By calculating the mean values of temperature and humidity within each cluster, a representative environmental stress combination for the cluster can be obtained, and the combination includes a temperature value and a humidity value. Thus, the setting of environmental stress conditions is no longer randomly selected but based on scientific analysis of historical meteorological data, ensuring that the selected environmental stress conditions can represent the common climate types in the expected usage area. It should be noted that during clustering, the temperature and humidity at the same time point in the historical meteorological data form a two-dimensional data point. Clustering analysis is performed on these two-dimensional data points. Specifically, each two-dimensional data point is equivalent to a two-dimensional coordinate, and a clustering algorithm is used to perform clustering analysis on a series of two-dimensional coordinate points.

[0038] Among them, in step A203, these typical environmental stress combinations are then used to set different channels of the environmental stress acceleration device, so that each channel simulates a typical environmental stress condition. Each parallel sample group is placed in these channels for an accelerated degradation experiment, and turbidity data is measured in real time in step A205 to provide data support for subsequent dispersion stability evaluation.

[0039] Among them, in step A204, the duration of the degradation process can be set according to actual needs, such as 7 days, 10 days, etc.

[0040] Among them, in step A205, after the experiment starts, the turbidity detection device automatically collects the turbidity values of each parallel sample at preset time intervals, such as every 5 minutes, 10 minutes, or 30 minutes. The collected data can be stored in the data memory.

[0041] Specifically, obtaining the historical meteorological data of the expected usage area through step A201 ensures the relevance of the subsequent analyzed environmental stress conditions to the actual application scenario. Step A202 is the core of the method. The application of the clustering algorithm realizes the extraction of a small number of representative typical environmental stress combinations from a large amount of historical meteorological data. Clustering analysis can identify the internal structure in the dataset, classify similar meteorological conditions, and represent each category with a typical environmental stress combination, avoiding the randomness in the selection of environmental stress conditions, ensuring that the selected environmental stress conditions can represent the common climate types in the expected usage area, and solving the problem of the lack of scientific basis in setting environmental stress conditions in traditional methods. Step A203 applies the extracted typical environmental stress combinations to the channel settings of the environmental stress acceleration device, enabling the experimental conditions to simulate the real environment and improving the practical guiding significance of the experimental results. Steps A204 and A205 conduct the deterioration experiment of the suspension dispersibility stability and the turbidity data measurement under the set typical environmental stress conditions, providing a data basis for the subsequent evaluation of the dispersibility stability. Through the above steps, the scientific selection of environmental stress conditions is achieved, making the dispersibility stability prediction method more suitable for the actual application environment, improving the accuracy and representativeness of the prediction, and providing more efficient technical support for the formulation optimization and quality control of herbicide suspensions.

[0042] In some possible implementation manners, step A202 includes: B1. For the historical meteorological data, perform clustering analysis using multiple clustering algorithms respectively to obtain multiple clustering results; the clustering algorithms include using the K-means clustering algorithm, hierarchical clustering algorithm, and DBSCAN clustering algorithm; B2. According to multiple evaluation indicators, evaluate the multiple clustering results and select the optimal clustering result; the evaluation indicators include the silhouette coefficient, Calinski-Harabasz index, and Davies-Bouldin index; B3. Calculate the average temperature and average humidity in each cluster of the optimal clustering result to obtain the typical environmental stress combination of the corresponding cluster.

[0043] Among them, in step B1, for the historical meteorological data, using multiple clustering algorithms is to explore the potential environmental stress combination patterns in the data from different perspectives. Specifically, the K-means clustering algorithm divides clusters based on distance and is suitable for discovering spherical clusters; the hierarchical clustering algorithm forms a hierarchical structure by gradually merging or splitting data points and can reveal the nested relationships between data points; the DBSCAN clustering algorithm identifies clusters based on density and can effectively identify clusters of any shape and noise points. By comprehensively applying these three algorithms, more comprehensive and robust clustering results can be obtained.

[0044] Among them, in step B2, after obtaining multiple clustering results, in order to select the optimal clustering result, multiple evaluation metrics are introduced to objectively evaluate the quality of the clustering results. The silhouette coefficient is used to evaluate the cohesion within clusters and the separation between clusters; the Calinski-Harabasz index evaluates the clustering effect through the ratio of the between-cluster dispersion to the within-cluster dispersion; the Davies-Bouldin index evaluates the clustering effect by calculating the between-cluster similarity and the within-cluster dispersion. These metrics can measure the quality of the clustering results from different aspects, and by comprehensively considering these metrics, the optimal clustering result can be selected more accurately. Among them, the calculation methods of the silhouette coefficient, the Calinski-Harabasz index, and the Davies-Bouldin index are existing technologies and will not be elaborated here.

[0045] Among them, in step B3, after determining the optimal clustering result, by calculating the average temperature and average humidity of the data points in each cluster, a typical environmental stress combination is obtained. The environmental stress combination extracted in this way can represent the environmental characteristics of the center of each cluster and is used for subsequent evaluation of the dispersion stability of the suspending agent, which can improve the environmental representativeness of the evaluation experiment.

[0046] Specifically, the purpose of this solution is to solve the problem of unclear clustering algorithm selection and clustering result evaluation criteria when using a clustering algorithm to extract a typical environmental stress combination. Through step B1, multiple clustering algorithms such as the K-means clustering algorithm, hierarchical clustering algorithm, and DBSCAN clustering algorithm are used to mine different clustering patterns from historical meteorological data, thereby obtaining multiple clustering results, ensuring the comprehensiveness of the clustering analysis. Then, through step B2, multiple evaluation metrics such as the silhouette coefficient, Calinski-Harabasz index, and Davies-Bouldin index are introduced to objectively evaluate the clustering results obtained by different clustering algorithms and select the clustering result with the best score, ensuring the quality of the clustering results. Finally, in step B3, based on the optimal clustering result, the average temperature and average humidity of each cluster are calculated to obtain a typical environmental stress combination, so that the extracted environmental stress combination can represent the typical environmental characteristics of the historical meteorological data. Through the above steps, this solution solves the problem of unclear algorithm selection and result evaluation criteria in the process of extracting the environmental stress combination, improves the accuracy and reliability of the extraction of the environmental stress combination, enables the subsequent evaluation experiment of the dispersion stability of the suspending agent to more effectively simulate real environmental conditions, and thus improves the accuracy of the evaluation results.

[0047] Preferably, step B1 may include: B101. For the historical meteorological data, standardize the temperature data and humidity data to eliminate the dimensional difference; B102. The elbow method is used to determine the optimal number of clusters for the K-means clustering algorithm. Based on the determined optimal number of clusters, the K-means clustering algorithm is used to perform clustering analysis on the standardized historical meteorological data to generate multiple clusters, obtaining the clustering result of the K-means algorithm; B103. For the hierarchical clustering algorithm, the Wardlinkage method is used to calculate the distance between clusters, construct a hierarchical clustering dendrogram, and according to the set threshold or the number of clusters, the dendrogram is segmented into multiple clusters, obtaining the clustering result of the hierarchical clustering algorithm; B104. For the DBSCAN clustering algorithm, the grid search method is used to determine the parameters of the neighborhood radius and the minimum number of samples. Based on the determined parameters of the neighborhood radius and the minimum number of samples, the DBSCAN clustering algorithm is used to perform clustering analysis on the historical meteorological data to generate multiple clusters, obtaining the clustering result of the DBSCAN algorithm.

[0048] Among them, in step B101, there are differences in the dimensions of the temperature data and the humidity data. Therefore, by performing standardized processing on the temperature data and the humidity data, such as using the Z-score standardization method or the Min-Max standardization method, the deviation of the clustering analysis result caused by the difference in the dimensions of the temperature and humidity data is eliminated.

[0049] Among them, in step B102, the elbow method is used to determine the optimal number of clusters for the K-means clustering algorithm. Specifically, the elbow method determines the optimal number of clusters by plotting the relationship between the number of clusters and the sum of squared errors, and selects the number of clusters where the rate of decrease in the sum of squared errors significantly slows down as the optimal number of clusters (the specific process is prior art and will not be elaborated here). Subsequently, based on the determined optimal number of clusters, the K-means clustering algorithm is used to perform clustering analysis on the standardized historical meteorological data, obtaining the clustering result of the K-means algorithm.

[0050] Among them, in step B103, the Wardlinkage method is used to calculate the distance between clusters in the hierarchical clustering algorithm. The Wardlinkage method aims to minimize the within-cluster variance and construct a hierarchical clustering dendrogram. The segmentation of the dendrogram is achieved by setting a threshold or the number of clusters, thereby obtaining the clustering result of the hierarchical clustering algorithm (the specific process is prior art and will not be elaborated here).

[0051] Among them, in step B104, the grid search method is used to determine the parameters of the neighborhood radius and the minimum number of samples for the DBSCAN clustering algorithm. The grid search method evaluates the clustering effect by traversing the set parameter combinations and selects the optimal parameter combination. Subsequently, based on the determined parameters of the neighborhood radius and the minimum number of samples, the DBSCAN clustering algorithm is used to perform clustering analysis on the historical meteorological data, obtaining the clustering result of the DBSCAN algorithm (the specific process is prior art and will not be elaborated here).

[0052] Specifically, in step B1, the historical meteorological data is first standardized through step B101, eliminating the interference of the dimension difference between temperature and humidity data volumes on the clustering analysis and ensuring the accuracy of the clustering analysis. Then, through steps B102, B103, and B104, the standardized historical meteorological data is clustered using three different clustering algorithms: the K-means clustering algorithm, the hierarchical clustering algorithm, and the DBSCAN clustering algorithm. Since different clustering algorithms have different clustering principles and applicable scenarios, using multiple clustering algorithms for clustering analysis can mine the potential distribution characteristics of historical meteorological data from multiple perspectives, improve the reliability of the clustering results, and avoid the problem of inaccurate clustering results caused by inappropriate selection of a single clustering algorithm. Further, in step B102, the elbow method is used to determine the optimal number of clusters for the K-means clustering algorithm, and in step B104, the grid search method is used to determine the neighborhood radius and minimum sample number parameters of the DBSCAN clustering algorithm. Determining the reasonable parameters of the clustering algorithm through parameter optimization methods can improve the accuracy of the clustering results, enabling the clustering results to more truly reflect the distribution characteristics of historical meteorological data and providing guarantee for the accuracy of subsequent environmental stress combination extraction.

[0053] Further, step B2 may include: B201. Calculate the silhouette coefficient, Calinski-Harabasz index, and Davies-Bouldin index for multiple clustering results respectively; B202. For each evaluation index, use normalization and reciprocal processing to map the values of each evaluation index to the [0, 1] interval to obtain the normalized evaluation index; B203. For each clustering result, perform weighted summation of the normalized silhouette coefficient, normalized Calinski-Harabasz index, and normalized Davies-Bouldin index to obtain a comprehensive score; B204. Select the clustering result with the highest comprehensive score as the optimal clustering result.

[0054] Among them, in step B201, for multiple clustering results obtained by multiple clustering algorithms, the silhouette coefficient, Calinski-Harabasz index, and Davies-Bouldin index are calculated respectively to quantify the quality of each clustering result. The calculation method of the silhouette coefficient can be that, first, calculate the average distance between each sample and other samples in its cluster, and the average distance between this sample and all samples in the nearest cluster, and then calculate the silhouette coefficient through these two distances. The closer the silhouette coefficient value is to 1, the better the clustering effect. The calculation method of the Calinski-Harabasz index can be to calculate the ratio of the between-cluster dispersion to the within-cluster dispersion. The higher the Calinski-Harabasz index value, the better the clustering effect. The calculation method of the Davies-Bouldin index can be to calculate the similarity between each cluster and its most similar cluster. The lower the Davies-Bouldin index value, the better the clustering effect.

[0055] Among them, in step B202, since the higher the values of the silhouette coefficient and Calinski-Harabasz index indicate the better the clustering effect, while the lower the value of the Davies-Bouldin index indicates the better the clustering effect. To unify the evaluation criteria, it is necessary to take the reciprocal of the Davies-Bouldin index so that the higher the value, the better the clustering effect. At the same time, since the value ranges of different evaluation indicators may vary, in order to make different evaluation indicators comparable, it is necessary to normalize the values of each evaluation indicator to the interval [0, 1]. The normalization method can be, for example, to adopt the min-max normalization method to linearly transform the original data to the interval [0, 1].

[0056] Among them, in step B203, for each clustering result, the normalized silhouette coefficient, the normalized Calinski-Harabasz index, and the Davies-Bouldin index after reciprocal processing are weighted and summed to obtain a comprehensive score. The way of weighted summation can be, for example, to assign the same or different weights to the three indicators, and then add the weighted values to obtain the comprehensive score.

[0057] Among them, in step B204, compare the comprehensive scores of each clustering result, and select the clustering result with the highest comprehensive score as the optimal clustering result for subsequent steps to be based on.

[0058] Specifically, a technical solution for selecting the optimal clustering result from multiple clustering results is provided herein. After obtaining multiple clustering results through step B1, in order to select the optimal clustering result therefrom, first, the silhouette coefficient, Calinski-Harabasz index, and Davies-Bouldin index of each clustering result are calculated through step B201. These indices quantify the quality of the clustering results from different perspectives. Then, through step B202, these evaluation metrics are normalized and reciprocated to make different metrics comparable and have a consistent evaluation direction. Next, in step B203, each evaluation metric is comprehensively considered through weighted summation to obtain the comprehensive score of each clustering result. The higher the comprehensive score, the better the quality of the clustering result. Finally, through step B204, the clustering result with the highest comprehensive score is selected as the optimal clustering result. Through the above steps, the optimal clustering result can be objectively and effectively selected from multiple clustering results, providing a more reliable basis for subsequent analysis, and further improving the accuracy of predicting the dispersion stability of the suspending agent.

[0059] In some embodiments, step A3 includes: A301. Extract the turbidity data of each parallel sample at different time points to form a turbidity time series; A302. For each turbidity time series, calculate the turbidity change rate using the sliding window method; A303. For each parallel sample, calculate the average turbidity change rate of the entire deterioration process according to the corresponding calculated result of the turbidity change rate, and convert it into a dispersion stability performance evaluation index.

[0060] Among them, in step A301, the turbidity data can be read from the data storage and arranged in chronological order to form a turbidity time series. Thus, the dynamic change process of the dispersion stability of the suspending agent can be comprehensively reflected, providing a data basis for subsequent calculation of the turbidity change rate and evaluation of the dispersion stability performance.

[0061] Among them, in step A302, the size and step length of the sliding window can be set. For example, the window size is set to 10 time points, and the step length is set to 1 time point. Then, starting from the starting moment of the turbidity time series, the window is slid according to the step length, and within each window, the turbidity change rate is calculated. The turbidity change rate can be calculated using various methods. For example, calculate the linear regression slope of the turbidity data within the window, or calculate the differential mean of the turbidity data within the window, or divide the difference in turbidity data between the end point and the starting point of the window by the window time length. Thus, the fluctuations of the turbidity time series can be effectively smoothed, the trend of turbidity change can be captured, and the change rate of the dispersion stability of the suspending agent can be accurately reflected.

[0062] Among them, in step A303, the average value of the turbidity change rates of all sliding windows during the entire deterioration process of each parallel sample can be calculated as the average turbidity change rate of the parallel sample. The dispersion stability performance evaluation index can be the average turbidity change rate itself, or a value obtained by mapping the average turbidity change rate to a specific interval. For example, the average turbidity change rate is normalized to 0 - 100 points, and a higher value indicates better dispersion stability performance. Thus, the overall level of the suspension agent's dispersion stability during the entire deterioration process can be comprehensively reflected and converted into a quantifiable evaluation index, facilitating the comparison and evaluation of the dispersion stability performance between different samples and under different conditions.

[0063] Specifically, in this solution, first, by extracting the turbidity time series through step A301, the dynamic change process of the suspension agent's dispersion stability is comprehensively tracked and described in a data - based manner. Then, by using the sliding window method to calculate the turbidity change rate through step A302, the data fluctuation is effectively smoothed, the trend of turbidity change is highlighted, and the accuracy of rate calculation is improved. Finally, by calculating the average turbidity change rate and converting it into a performance evaluation index through step A303, the quantitative evaluation of the dispersion stability performance is achieved, facilitating performance comparison and evaluation. Through the above steps, a clear, specific, and operable dispersion stability evaluation method is provided, making the technical solution more practical and effective.

[0064] Through the above - mentioned technical solution, this application makes the calculation method of the turbidity change rate clearer and the evaluation method of the dispersion stability performance more specific, improving the accuracy and reliability of the dispersion stability evaluation results. It provides more effective data support for subsequent optimization of the suspension agent to be tested based on the dispersion stability performance evaluation results and for analyzing the influence level of each environmental factor on the dispersion stability.

[0065] In some preferred embodiments, step A303 includes: For each parallel sample, extract the turbidity change rate time series and divide it into multiple time windows according to the time sequence; For each time window, calculate the average weighted value of the turbidity change rate; the farther the time window is from the initial moment of the deterioration process, the smaller the corresponding weight coefficient; For each parallel sample, accumulate the average weighted values of the turbidity change rates of each time window as the dispersion stability performance evaluation index.

[0066] Among them, the time window refers to multiple time periods into which the time series of the turbidity change rate is divided, and it is not the same window as the sliding window mentioned above. Among them, the time window can be divided in an equally spaced or unequally spaced manner. For example, a time window can be set every 2 hours, or according to the characteristics of the turbidity change rate, a smaller time window can be set in the stage with a faster turbidity change rate, and a larger time window can be set in the stage with a slower turbidity change rate.

[0067] Among them, the farther the time window is from the initial moment of the deterioration process, the smaller the corresponding weight coefficient is. This is to highlight the role of early deterioration information in the comprehensive evaluation. Different weight coefficients are assigned to different time windows because the deterioration speed is faster in the early stage. The setting method of the weight coefficient can adopt linear decrease, exponential decrease, or piecewise decrease, etc. For example, the weight coefficient of the first time window can be set to 1, the weight coefficient of the second time window can be set to 0.9, and so on. The weight coefficient decreases linearly as the distance of the time window from the initial moment increases. Another example is that the weight coefficient can be set to decrease exponentially as the distance of the time window from the initial moment increases. For example, the weight coefficient is the nth power of 0.98, where n is the serial number of the time window from the initial moment.

[0068] Among them, for each time window, the average weighted value of the turbidity change rate is calculated to obtain a value that can better represent the overall level of the turbidity change rate within the time window. The calculation method of the average weighted value can be to first calculate the average value of the turbidity change rate within the time window, and then multiply this average value by the weight coefficient corresponding to this time window.

[0069] Among them, accumulating the average weighted values of the turbidity change rates of each time window means accumulating the average weighted values calculated for all time windows to obtain an overall evaluation index for comprehensively evaluating the dispersion stability performance of the suspending agent.

[0070] Specifically, this solution aims to more accurately evaluate the dispersion stability performance. Instead of simply calculating the average turbidity change rate of the entire deterioration process, the concepts of time window and weighted average are introduced to more accurately reflect the dynamic characteristics of the dispersion stability over time.

[0071] First, the time series of the turbidity change rate is divided into multiple time windows in chronological order. This can decompose the entire deterioration process into multiple stages for analysis and capture the characteristics of the turbidity change in different time periods.

[0072] Secondly, for each time window, calculate the average weighted value of the turbidity change rate and set the weight coefficient such that the farther the time window is from the initial moment of the deterioration process, the smaller the corresponding weight coefficient. The significance of this weighting method lies in that in the early stage of the deterioration of the suspension dispersion stability, the turbidity change rate may be more sensitive and can better reflect the potential instability factors of the dispersion system. Therefore, assigning a higher weight to the early time window can highlight the role of early deterioration information in the comprehensive evaluation. As the deterioration process progresses, the turbidity change rate may tend to flatten out, and the impact of later changes on the overall stability weakens relatively. Therefore, it is reasonable to reduce the weight of the later time window.

[0073] Finally, the average weighted values of the turbidity change rates of each time window of each parallel sample are accumulated to obtain a comprehensive evaluation index for the dispersion stability performance. This index comprehensively considers the turbidity change conditions in different time periods and highlights the impact of early deterioration, thereby being able to more comprehensively and accurately evaluate the dispersion stability performance of the suspension, overcoming the deficiency that the simple average method may ignore important early deterioration information, making the evaluation result more in line with the actual situation, and providing a more reliable basis for subsequent formulation optimization and quality control.

[0074] Through the above technical solution, the present application can more accurately evaluate the dispersion stability performance of the suspension. Especially when the turbidity change rate tends to flatten out in the later stage of the deterioration process, the impact of the early turbidity change rate on the dispersion stability is effectively highlighted. Thus, the evaluation result can be more in line with the actual situation and provide a more reliable basis for subsequent formulation optimization and quality control.

[0075] In some embodiments, step A4 includes: A401. Obtain the historical meteorological data of the expected use area, and calculate the occurrence probability of each environmental stress condition in the expected use area according to the historical meteorological data; A402. Convert the occurrence probability of each environmental stress condition in the expected use area into a weight coefficient; A403. Calculate the weighted average dispersion stability performance based on the dispersion stability performance of each parallel sample and the corresponding weight coefficient, and use it as the comprehensive dispersion stability performance of the suspension to be tested.

[0076] Among them, in step A401, the way to obtain the historical meteorological data can refer to step A201. Specifically, for the calculation of the occurrence probability of the environmental stress condition, it can be to count the frequencies of various environmental stress conditions in the historical meteorological data, and then divide the frequency of each environmental stress condition by the total number of observations to obtain the occurrence probability of this environmental stress condition in the expected use area.

[0077] Among them, in step A402, the conversion of the occurrence probability of environmental stress conditions to weight coefficients can be carried out by means of linear mapping. For example, the higher the occurrence probability, the larger the weight coefficient, and vice versa. Nonlinear mapping can also be used to emphasize or weaken the influence of certain specific environmental stress conditions.

[0078] Among them, in step A403, the calculation of the weighted average dispersion stability performance is specifically to multiply the dispersion stability performance of parallel samples under each environmental stress condition by the corresponding weight coefficient, and then add up all the products to obtain the weighted average dispersion stability performance, which is used as the comprehensive dispersion stability performance of the suspension agent to be tested.

[0079] Specifically, for the problem of comprehensive dispersion stability performance evaluation, a weighted average method based on the occurrence probability of environmental stress conditions is proposed. In step A401, by analyzing the historical meteorological data of the expected use area, the possibility of different environmental stress conditions occurring in this area is quantified, which provides an environmental background basis for subsequent comprehensive evaluation. In step A402, these occurrence probabilities are converted into weight coefficients, so that in the comprehensive evaluation, the environmental stress conditions that occur more frequently have a greater impact on the final result. Step A403 then uses these weight coefficients to perform a weighted average calculation on the dispersion stability performance of each parallel sample under different environmental stress conditions to obtain the comprehensive dispersion stability performance of the suspension agent to be tested. Thus, the evaluation result is closer to the true performance of the suspension agent to be tested in the actual application scenario, improving the accuracy and practicality of the evaluation.

[0080] In some embodiments, in step A5, first, the dispersion stability performance of each parallel sample is used as the dependent variable, and the corresponding simulated irrigation water hardness, temperature, and humidity are used as independent variables to construct a regression model. The regression model can be a multiple linear regression model or a nonlinear regression model. Interaction terms can be introduced into the regression model, such as the interaction term between irrigation water hardness and temperature, the interaction term between temperature and humidity, etc., to analyze the influence of the interaction between environmental factors on the dispersion stability performance. Based on the constructed regression model, regression analysis is carried out to obtain the regression coefficients of each environmental factor and interaction term. Among them, the absolute value of the regression coefficient can represent the influence degree of the corresponding environmental factor or interaction term on the dispersion stability performance. The larger the absolute value, the higher the influence level. Finally, according to the size of the regression coefficient, the influence level of each environmental factor on the dispersion stability of the suspension agent to be tested is determined (for example, different influence levels are preset in advance, and a corresponding regression coefficient range is set for each influence level. According to the regression coefficient range into which the calculated regression coefficient falls, the corresponding influence level is determined), and whether there is an interaction and nonlinear influence between each environmental factor is analyzed.

[0081] Reference Figure 2, this application also proposes an optimization system for predicting the dispersion stability of a herbicide suspension concentrate, which includes: An environmental stress acceleration device 1, which has multiple channels, and the environmental stress conditions of each channel can be independently adjusted. Each channel is used to accelerate the deterioration of the dispersion stability of the suspension concentrate for different parallel sample groups under different environmental stress conditions. The environmental stress conditions include temperature conditions and humidity conditions. The parallel sample groups are obtained in the following way: Dilute the suspension concentrate to be tested with simulated irrigation water of multiple different hardnesses to form multiple initial suspension concentrate samples, and divide each initial suspension concentrate sample into multiple parallel samples to form multiple parallel sample groups. Each parallel sample group contains parallel samples corresponding to simulated irrigation water of all hardnesses; A turbidity detection device 2, which is used to measure the turbidity data of each parallel sample in real time during the deterioration process; A host computer 3, which is used to perform the following steps: According to the measured turbidity data, calculate the turbidity change rate of each parallel sample to evaluate the dispersion stability performance of each parallel sample; According to the dispersion stability performance of each parallel sample, evaluate the comprehensive dispersion stability performance of the suspension concentrate to be tested; According to the dispersion stability performance of each parallel sample, the hardness of the corresponding simulated irrigation water, and the environmental stress conditions, analyze the influence level of each environmental factor on the dispersion stability of the suspension concentrate to be tested. The environmental factors include the hardness of irrigation water, temperature, and humidity; Output an evaluation report including the dispersion stability performance of each parallel sample, the comprehensive dispersion stability performance, and the influence level of each environmental factor on the dispersion stability of the suspension concentrate to be tested.

[0082] Among them, the environmental stress acceleration device 1 is configured to have a plurality of independent channels, which can be independently adjusted in terms of temperature conditions and humidity conditions, so as to conduct a dispersion stability accelerated degradation experiment on a parallel sample group under different environmental stresses. The parallel sample group is prepared as follows: First, the suspension agent to be tested is diluted with simulated irrigation water of various hardnesses to form a plurality of initial suspension agent samples; subsequently, each initial suspension agent sample is divided into a plurality of parts and placed in different sample containers to form a plurality of parallel samples, and finally a plurality of parallel sample groups are formed. Each parallel sample group contains parallel samples corresponding to all simulated irrigation waters of different hardnesses, so as to simulate the influence of water quality hardness changes on the suspension agent stability in actual applications. The turbidity detection device 2 is set to collect the turbidity data of each parallel sample in real time during the suspension agent degradation experiment to monitor the change of the dispersion state of the suspension agent. The upper computer 3 serves as a data processing and analysis center, and its function is realized as follows: First, based on the data collected by the turbidity detection device 2, calculate the turbidity change rate of each parallel sample during the degradation process, and use this rate as an index to evaluate the dispersion stability performance; then, comprehensively evaluate the dispersion stability performance of each parallel sample to make an overall evaluation of the dispersion stability of the suspension agent to be tested; after that, combine the dispersion stability performance of each parallel sample and the experimental conditions (simulated irrigation water hardness, temperature and humidity) to analyze the influence degree of these environmental factors on the suspension agent dispersion stability, and determine the key influencing factors; finally, the system generates an evaluation report, and the report content covers the dispersion stability performance of each parallel sample, the comprehensive dispersion stability performance of the suspension agent to be tested, and the influence level of each environmental factor, providing data support for subsequent formulation optimization.

[0083] Specifically, the working principle of this system lies in simulating the actual field environmental conditions, accelerating the dispersion stability degradation process of the herbicide suspension agent, and conducting a rapid evaluation. The environmental stress acceleration device 1 provides a controllable temperature and humidity environment to simulate different climate conditions; by diluting the suspension agent with simulated irrigation water of various hardnesses, the water quality differences in different regions are simulated. During the experiment, the turbidity detection device 2 monitors the dispersion state of the suspension agent in real time, and the turbidity change rate reflects the quality of the suspension agent dispersion stability. The upper computer 3 quantifies the dispersion stability performance and evaluates the influence of different environmental factors by analyzing the turbidity data. Thus, this system can quickly obtain the dispersion stability data of the herbicide suspension agent under various environmental conditions, overcome the defect of the long time-consuming traditional long-term storage experiment, and provide a technical means for small pesticide enterprises to quickly optimize the formulation and ensure product quality. By analyzing the influence level of environmental factors, the key factors affecting the suspension agent stability can be determined, providing a direction for formulation improvement. The final evaluation report provides a basis for enterprise decision-making and helps to enhance product competitiveness.

[0084] Among them, the environmental stress acceleration device 1 can adopt a multi-channel constant temperature and humidity chamber, the turbidity detection device 2 can adopt an on-line turbidimeter, and the upper computer 3 can be a device with computing power such as a personal computer, an industrial control computer, a server, etc.

[0085] The above are only embodiments of the present application and are not used to limit the protection scope of the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for predicting and optimizing the dispersion stability of a herbicide suspension concentrate, characterized in that, The method includes: A1. Dilute the suspending agent to be tested with simulated irrigation water of various different hardness levels to form multiple initial suspending agent samples, and divide each initial suspending agent sample into multiple parallel samples to form multiple groups of parallel samples; each group of parallel samples contains parallel samples corresponding to simulated irrigation water of all hardness levels; A2. Place each group of parallel samples under different environmental stress conditions to accelerate the deterioration of the dispersion stability of the suspending agent, and during the deterioration process, measure the turbidity data of each parallel sample in real time; the environmental stress conditions include temperature conditions and humidity conditions; A3. Calculate the turbidity change rate of each parallel sample based on the measured turbidity data to evaluate the dispersion stability performance of each parallel sample; A4. Evaluate the comprehensive dispersion stability performance of the suspending agent to be tested based on the dispersion stability performance of each parallel sample; A5. Analyze the influence level of each environmental factor on the dispersion stability of the suspending agent to be tested based on the dispersion stability performance of each parallel sample and the corresponding simulated irrigation water hardness and environmental stress conditions; the environmental factors include irrigation water hardness, temperature, and humidity; A6. Output an evaluation report including the dispersion stability performance of each parallel sample, the comprehensive dispersion stability performance of the suspending agent to be tested, and the influence level of each environmental factor on the dispersion stability of the suspending agent to be tested.

2. The method for predicting and optimizing the dispersion stability of a herbicide suspension concentrate according to claim 1, wherein Step A1 includes: A101. Configure simulated irrigation water of various different hardness levels according to the common water quality range of the expected usage area; A102. Dilute the suspending agent to be tested to the target concentration respectively according to the configured simulated irrigation water of different hardness levels to obtain multiple initial suspending agent samples; A103. Measure each initial suspending agent sample and dispense it into multiple sample containers, with each sample container corresponding to a parallel sample, to form multiple groups of parallel samples; each group of parallel samples contains parallel samples corresponding to simulated irrigation water of all hardness levels.

3. A method for predicting and optimizing the dispersion stability of a herbicide suspension concentrate, as claimed in claim 1, wherein Step A2 includes: A201. Obtain the historical meteorological data of the expected usage area, where the historical meteorological data includes temperature data and humidity data; A202. Use a clustering algorithm to extract multiple typical environmental stress combinations from the historical meteorological data; each typical environmental stress combination includes a temperature data and a humidity data; A203. Set the temperature and humidity of each channel of the environmental stress acceleration device so that the environmental stress conditions of each channel correspond to a typical environmental stress combination; A204. Place each group of parallel samples into each channel of the environmental stress acceleration device respectively to accelerate the deterioration of the dispersion stability of the suspending agent; A205. During the deterioration process, measure the turbidity data of each parallel sample in real time.

4. The method for predicting and optimizing the dispersion stability of a herbicide suspension concentrate according to claim 3, wherein Step A202 includes: B1. Conduct clustering analysis on the historical meteorological data using various clustering algorithms respectively to obtain multiple clustering results; the clustering algorithms include using the K-means clustering algorithm, hierarchical clustering algorithm, and DBSCAN clustering algorithm; B2. Evaluate the multiple clustering results according to multiple evaluation indicators and select the optimal clustering result; the evaluation indicators include the silhouette coefficient, Calinski-Harabasz index, and Davies-Bouldin index; B3. Calculate the average temperature and average humidity in each cluster of the optimal clustering result to obtain the typical environmental stress combination for the corresponding cluster.

5. A method for predicting and optimizing the dispersion stability of a herbicide suspension concentrate, characterized in that, Step B1 includes: B101. For the historical meteorological data, perform standardization processing on the temperature data and humidity data to eliminate the dimension difference; B102. Use the elbow method to determine the optimal number of clusters for the K-means clustering algorithm, and based on the determined optimal number of clusters, use the K-means clustering algorithm to perform clustering analysis on the standardized historical meteorological data to generate multiple clusters, obtaining the clustering result of the K-means algorithm; B103. For the hierarchical clustering algorithm, use the Ward linkage method to calculate the distance between clusters, construct a hierarchical clustering dendrogram, and according to the set threshold or number of clusters, divide the dendrogram into multiple clusters, obtaining the clustering result of the hierarchical clustering algorithm; B104. For the DBSCAN clustering algorithm, use the grid search method to determine the parameters of the neighborhood radius and the minimum number of samples, and based on the determined neighborhood radius and minimum number of samples, use the DBSCAN clustering algorithm to perform clustering analysis on the historical meteorological data to generate multiple clusters, obtaining the clustering result of the DBSCAN algorithm.

6. The method for predicting and optimizing the dispersion stability of a herbicide suspension concentrate according to claim 4, wherein, Step B2 includes: B201. For multiple clustering results, calculate the silhouette coefficient, Calinski-Harabasz index, and Davies-Bouldin index respectively; B202. For each evaluation index, perform normalization and reciprocal processing to map the values of each evaluation index to the interval [0, 1], obtaining the normalized evaluation index; B203. For each clustering result, perform weighted summation of the normalized silhouette coefficient, normalized Calinski-Harabasz index, and normalized Davies-Bouldin index to obtain a comprehensive score; B204. Select the clustering result with the highest comprehensive score as the optimal clustering result.

7. A method for predicting and optimizing the dispersion stability of a herbicide suspension concentrate, characterized in that, Step A3 includes: A301. Extract the turbidity data of each parallel sample at different time points to form a turbidity time series; A302. For each turbidity time series, use the sliding window method to calculate the turbidity change rate; A303. For each parallel sample, calculate the average turbidity change rate of the entire deterioration process according to the corresponding turbidity change rate calculation result, and convert it into a dispersion stability performance evaluation index.

8. A method for predicting and optimizing the dispersion stability of a herbicide suspension concentrate, characterized in that, Step A303 includes: For each parallel sample, extract the turbidity change rate time series and divide it into multiple time windows according to the time sequence; For each time window, calculate the average weighted value of the turbidity change rate; the farther the time window is from the initial moment of the deterioration process, the smaller the corresponding weight coefficient; For each parallel sample, accumulate the average weighted values of the turbidity change rate of each time window as the dispersion stability performance evaluation index.

9. The method for predicting and optimizing the dispersion stability of a herbicide suspension concentrate according to claim 1, characterized in that, Step A4 includes: A401. Obtain the historical meteorological data of the expected use area, and calculate the occurrence probability of each environmental stress condition in the expected use area according to the historical meteorological data; A402. Convert the occurrence probability of each environmental stress condition in the expected use area into a weight coefficient; A403. Calculate the weighted average dispersion stability performance based on the dispersion stability performance of each parallel sample and the corresponding weight coefficient, and use it as the comprehensive dispersion stability performance of the suspending agent to be tested.

10. A system for predicting and optimizing the dispersion stability of a herbicide suspension concentrate, characterized in that, The system includes: An environmental stress acceleration device with multiple channels, the environmental stress conditions of each channel being independently adjustable. Each of the channels is used to accelerate the deterioration of the dispersion stability of the suspending agent for different parallel sample groups under different environmental stress conditions. The environmental stress conditions include temperature conditions and humidity conditions. The parallel sample groups are obtained by the following method: Dilute the suspending agent to be tested with simulated irrigation water of multiple different hardness levels to form multiple initial suspending agent samples, and divide each initial suspending agent sample into multiple parallel samples to form multiple parallel sample groups. Each parallel sample group contains parallel samples corresponding to simulated irrigation water of all hardness levels. A turbidity detection device for measuring the turbidity data of each parallel sample in real time during the deterioration process. A host computer for performing the following steps: Calculate the turbidity change rate of each parallel sample based on the measured turbidity data to evaluate the dispersion stability performance of each parallel sample. Evaluate the comprehensive dispersion stability performance of the suspending agent to be tested based on the dispersion stability performance of each parallel sample. Analyze the influence level of each environmental factor on the dispersion stability of the suspending agent to be tested based on the dispersion stability performance of each parallel sample and the corresponding simulated irrigation water hardness and environmental stress conditions. The environmental factors include irrigation water hardness, temperature, and humidity. Output an evaluation report including the dispersion stability performance of each parallel sample, the comprehensive dispersion stability performance, and the influence level of each environmental factor on the dispersion stability of the suspending agent to be tested.

Images