A method and system for optimizing daylily field management based on big data

Abstract

The present invention discloses a daylily field management optimization method and system based on big data, comprising: acquiring environmental monitoring data of a daylily field, mining the environmental monitoring data through big data analysis to obtain control data, and guiding the environmental control of the daylily according to the control data; predicting the environmental monitoring data through a machine learning algorithm to obtain expected growth conditions of the daylily; acquiring meteorological data and image data of the daylily, obtaining disease data and growth conditions according to the image data, predicting the growth stage and yield of the daylily according to the meteorological data, environmental monitoring data, disease data and growth conditions, identifying and optimizing according to the growth stage, yield, environmental monitoring data and disease data through a rule engine, and obtaining a control plan to further regulate the field management of the daylily.

Description

Technical Field

[0001] The present invention belongs to the technical field of data management, and in particular relates to a daylily field management optimization method and system based on big data. Background Art

[0002] A key technical challenge in daylily field management is how to achieve real-time monitoring and data collection of key parameters during the plant's growth process. While various sensors currently offer real-time monitoring of parameters such as temperature, humidity, and light intensity, practical applications still face numerous challenges. First, the complex and dynamic nature of daylily's growth environment, with varying environmental requirements at different stages, makes it difficult for a single sensor type to comprehensively monitor all parameters. Second, the harsh field environment of daylily requires sensors with strong anti-interference capabilities and environmental adaptability, otherwise data distortion and equipment damage are likely to occur. Furthermore, the long growth cycle of daylily requires sensors to operate stably for extended periods, placing high demands on their power supply and data transmission. Finally, the effective integration and management of massive amounts of monitoring data presents a significant challenge, requiring the establishment of a comprehensive data collection, transmission, and storage system while ensuring data accuracy and security. Addressing these challenges requires the comprehensive application of multiple technical approaches, including optimizing sensor layout, improving sensor environmental adaptability and stability, and establishing an efficient data collection and management platform. This is crucial for achieving precise and intelligent daylily field management. Summary of the Invention

[0003] In order to solve the above technical problems, the present invention proposes a daylily field management optimization method and system based on big data to solve the problems existing in the above-mentioned prior art.

[0004] To achieve the above objectives, the present invention provides a method and system for optimizing daylily field management based on big data, comprising:

[0005] Acquire environmental monitoring data of a daylily field, mine the environmental monitoring data through big data analysis to obtain control data, and guide environmental control of the daylily according to the control data;

[0006] Predicting the environmental monitoring data using a machine learning algorithm to obtain expected growth conditions of the daylily;

[0007] Acquire meteorological data and image data of the daylily, obtain disease data and growth status based on the image data, predict the growth stage and yield of the daylily based on the meteorological data, environmental monitoring data, disease data and growth status, and use a rule engine to identify and optimize based on the growth stage, yield, environmental monitoring data and disease data to obtain a control plan to further regulate the field management of the daylily.

[0008] Optionally, the environmental monitoring data includes soil temperature, humidity, nitrogen, phosphorus and potassium content, light intensity, air temperature, humidity and carbon dioxide concentration.

[0009] Optionally, before mining the environmental monitoring data, the following steps may be further performed:

[0010] The environmental monitoring data is preprocessed, wherein the preprocessing includes noise removal and data cleaning and filtering.

[0011] Optionally, before mining the environmental monitoring data, the following steps may be further performed:

[0012] The environmental monitoring data is distributedly stored in a database, wherein the environmental monitoring data is classified by a machine learning method, and the environmental monitoring data is stored according to the classification results.

[0013] Optionally, the process of mining the environmental monitoring data includes:

[0014] Obtaining the growth status corresponding to the environmental monitoring data; training the decision tree model using the environmental monitoring data and the growth status to obtain a trained decision tree model, and performing analysis based on the trained decision tree model to obtain key environmental parameters;

[0015] The machine learning model is trained according to the key environmental parameters and the corresponding growth status, the key environmental parameters are iteratively adjusted through the optimization algorithm, and the key environmental parameters in the iterative adjustment are predicted by the trained machine learning model to obtain the growth status score as the fitness, wherein the key environmental parameters are iteratively adjusted according to the fitness until the stopping condition is reached, and the optimal growth condition combination of the key environmental parameters, i.e., the control data, is obtained.

[0016] Optionally, the process of predicting the environmental monitoring data includes:

[0017] The environmental monitoring data is predicted by a support vector machine to obtain the daylily growth data, wherein the daylily growth data includes plant height, number of leaves, and number of flower buds. The daylily growth data is predicted by a random forest model to obtain the expected growth of the daylily.

[0018] Optionally, the process of obtaining disease data and growth status includes:

[0019] The image data is identified by an image recognition method to obtain disease data and growth status.

[0020] Optionally, the process of obtaining the growth stages and yield of daylily includes:

[0021] The meteorological data, environmental monitoring data, disease data and growth conditions are processed by a machine learning model to obtain the growth stage and yield of the daylily.

[0022] Optionally, after obtaining the control plan, the method further includes visually displaying the environmental monitoring data, control data, expected growth conditions of the daylily, and the control plan through a visual interface.

[0023] On the other hand, the present invention also provides a daylily field management optimization system based on big data, which is used in the above method.

[0024] Compared with the prior art, the present invention has the following advantages and technical effects:

[0025] The present invention discloses a method for optimizing daylily field management based on big data. The present invention uses multiple types of sensors to comprehensively monitor the growth environment of daylily, integrates multi-source heterogeneous data through sensor fusion technology, and realizes effective integration of environmental data. A data management system is built on a cloud platform to pre-process and perform big data analysis on massive monitoring data, and to explore environmental laws and growth trends. Using machine learning algorithms such as support vector machines and random forests, environmental monitoring data and growth data are modeled and analyzed to train a daylily growth status prediction model. Based on this model, intelligent prediction and decision optimization of the growth process are realized, providing a basis for scientific planting management. The present invention also develops a growth supervision visualization system to intuitively display the real-time dynamics of the daylily growth environment and status, and presents the prediction results in the form of charts, providing farmers with data query and decision support functions. The present invention realizes intelligent monitoring of the daylily growth environment, accurate prediction of growth status and decision support for scientific planting management, effectively improving the scientific and intelligent level of daylily planting. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of this application. The exemplary embodiments and descriptions of this application are intended to explain this application and do not constitute an improper limitation on this application. In the accompanying drawings:

[0027] Figure 1 This is a flow chart of a daylily field management optimization method based on big data according to an embodiment of the present invention;

[0028] Figure 2 Another flow chart of the daylily field management optimization method according to an embodiment of the present invention;

[0029] Figure 3 This is a partial flow chart of the daylily field management optimization method according to an embodiment of the present invention;

[0030] Figure 4This is a structural diagram of the daylily field management optimization system based on big data in an embodiment of the present invention. DETAILED DESCRIPTION

[0031] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0032] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0033] like Figure 1-4 In this embodiment, a method and system for optimizing daylily field management based on big data may specifically include:

[0034] Step S101: Based on the growth environment of the daylily, various types of sensors are used to conduct comprehensive monitoring of temperature, humidity, light intensity, and soil moisture to obtain comprehensive environmental monitoring data. This includes:

[0035] According to the characteristics of the daylily's growing environment, select appropriate temperature and humidity, light intensity, and soil moisture sensors and arrange them reasonably to ensure that the sensors can fully collect environmental data. The sensors collect real-time temperature and humidity, light intensity, and soil moisture data of the daylily's growing environment;

[0036] Specifically, the design of a daylily growth environment monitoring system requires consideration of multiple aspects. The first is sensor selection and placement. The DHT22 temperature and humidity sensor, with its ±0.5°C and ±2% RH accuracy, is suitable for daylily growth environments. The BH1750 light intensity sensor, with a measurement range of 1-65535 lx, meets the needs of daylily at different growth stages. A capacitive soil moisture sensor can be used to avoid electrode corrosion. Environmental parameter data primarily includes soil temperature, humidity, nitrogen, phosphorus, and potassium content, light intensity, air temperature, humidity, and carbon dioxide concentration. Sensors for these factors can be positioned and configured based on sensor type.

[0037] These sensors should be evenly distributed throughout the daylily planting area to ensure representative data. Wireless sensor network technology can be used for data collection and transmission. The ZigBee protocol is low-power and low-cost, making it suitable for large-scale deployment. Each sensor node regularly collects environmental data and transmits it to a gateway via the ZigBee network. The gateway then uploads the data to a cloud data processing center via a 4G or NB-IoT network.

[0038] Step S102: Integrate and collaboratively process multi-source heterogeneous monitoring data from different sensors through sensor fusion technology to achieve effective data integration and obtain accurate environmental monitoring results. This includes:

[0039] The data processing center obtains multi-source heterogeneous monitoring data from different sensors, pre-processes data of different types and formats, unifies data formats and units, and eliminates redundant and noisy data.

[0040] According to environmental monitoring indicators and evaluation standards, the pre-processed data is analyzed and evaluated to determine whether the environmental quality meets the standards. If not, the early warning mechanism is triggered and an alarm message is issued.

[0041] Environmental monitoring results are visualized in charts, reports, and other formats for easy viewing and analysis. The resulting data is stored in a database to form historical monitoring data. Sensors are regularly calibrated and maintained to ensure the accuracy and stability of data collection. Fusion algorithms and evaluation models are optimized and updated to improve the system's adaptability and intelligence.

[0042] Specifically, multi-source heterogeneous data fusion is a key technology in environmental monitoring systems. For example, in monitoring the growth environment of daylilies, the system might simultaneously collect data on temperature, humidity, light intensity, and soil moisture. This data comes from different types of sensors, and the formats and units may differ. For example, temperature might be expressed in degrees Celsius or Fahrenheit, while humidity might be relative or absolute. Therefore, data preprocessing is necessary to unify the formats and units.

[0043] Environmental quality evaluation is an important part of data analysis. According to the growth requirements of daylily, the appropriate temperature range (such as 18-25℃), light intensity range (such as 3000-5000lux) and soil moisture range (such as 20%-30%) can be set. If the monitoring data exceeds these ranges, the system will trigger an early warning mechanism to remind farmers to take corresponding measures, such as adjusting irrigation or shading. Visual display is crucial for users to understand and analyze data. Line graphs can be used to show the changing trends of temperature, light and soil moisture, dashboards can be used to display the current environmental conditions, and heat maps can be used to show the temperature distribution in the greenhouse. These intuitive charts can help farmers quickly grasp the overall situation of the daylily growth environment.

[0044] Regular sensor calibration and maintenance are crucial for ensuring data quality. For example, temperature sensors can be calibrated monthly using a standard thermometer to ensure measurement accuracy. Furthermore, the system's fusion algorithms and evaluation models require continuous optimization. Machine learning techniques can be used to train models based on historical data, improving the system's accuracy in assessing the growing environment of daylilies. Through this series of data processing and analysis steps, the environmental monitoring system can provide comprehensive and accurate environmental information for daylily cultivation, helping farmers promptly identify and resolve potential issues, thereby improving yield and quality. This intelligent environmental monitoring approach is applicable not only to daylilies but can also be extended to the cultivation and management of other crops, providing strong support for the development of modern agriculture.

[0045] Step S103: Build a data management system on the cloud computing platform to perform pre-processing operations such as classification, storage, cleaning, and filtering on the massive monitoring data. Use big data analysis technology to mine environmental patterns and growth trends from the data, providing data support for optimizing the growth of daylilies. This includes:

[0046] A distributed data management system is built on a cloud computing platform, and big data frameworks such as Hadoop are used to distribute and store massive amounts of monitoring data. Based on pre-set classification rules, a decision tree algorithm is used to classify monitoring data in multiple dimensions, storing different types of data in corresponding databases.

[0047] Obtain classified monitoring data, remove noise data and outliers through data cleaning algorithms, and filter the data according to predefined filtering conditions to obtain a high-quality data set.

[0048] For the preprocessed data set, the association rule mining algorithm was used to analyze the association rules between environmental factors and the growth status of daylily, and the key environmental factors affecting the growth of daylily were obtained.

[0049] The time series analysis algorithm is used to predict the trend of daylily growth data, and the growth trend of daylily in the future is predicted based on historical data.

[0050] The discovered environmental patterns and predicted growth trends are visualized, providing agricultural experts with intuitive data support and assisting in developing strategies for optimizing daylily growth. Based on these patterns and growth trends, a machine learning algorithm was used to develop a daylily growth optimization model. By simulating growth under different environmental conditions, the optimal growth conditions were predicted to guide actual planting.

[0051] Specifically, a distributed data management system can be built through cloud computing platforms. For example, public cloud platforms such as Amazon Cloud or Alibaba Cloud can be utilized, or private cloud platforms can be built based on actual needs. This provides on-demand computing and storage resources, enabling flexible data management and efficient utilization. The core of a distributed data management system is to store massive amounts of data in a distributed manner across multiple nodes and implement unified management and scheduling. This approach improves data reliability and availability, avoids the risk of data loss due to single points of failure, and improves data processing efficiency. Distributed storage of massive monitoring data is achieved using big data frameworks such as Hadoop. Hadoop is an open-source distributed system infrastructure whose core components include the Distributed File System (HDFS) and the distributed computing framework (MapReduce). HDFS divides large files into multiple data blocks and stores them on different nodes, enabling distributed data storage and redundant backup. MapReduce decomposes computational tasks into multiple subtasks, distributes them to different nodes for parallel execution, and aggregates the results, achieving efficient distributed computing. For example, data on the growth environment of daylilies from different years and regions can be stored in different HDFS data blocks, facilitating subsequent query and analysis. Based on preset classification rules, a decision tree algorithm is used to perform multi-dimensional classification of monitoring data, storing different types of data in corresponding databases. A decision tree is a commonly used classification algorithm that makes classification decisions by constructing a tree-like structure. For example, monitoring data can be classified based on its source (such as soil sensors, meteorological sensors, etc.), monitoring indicators (such as temperature, humidity, light, etc.), and time range, with soil temperature data, air humidity data, light intensity data, etc. stored in different database tables. This improves data organization and query efficiency.

[0052] After obtaining classified monitoring data, a data cleaning algorithm is used to remove noise and outliers. The data is then filtered according to predefined filter conditions to obtain a high-quality data set. Data cleaning is an important step in data preprocessing, aiming to remove errors, incompleteness, and inconsistencies in the data.

[0053] For example, due to sensor failure or environmental interference, monitoring data may contain outliers, such as temperature data that suddenly drops to an extremely high or low value. These outliers need to be identified and removed. Filtering conditions can be set according to actual needs, for example, retaining only data within a specific time range or data from a specific region. Data cleaning and filtering can improve data accuracy and reliability, providing a high-quality data foundation for subsequent analysis and mining. An association rule mining algorithm is applied to the preprocessed dataset to analyze the association rules between environmental factors and daylily growth conditions, identifying the key environmental factors affecting daylily growth. Association rule mining is a commonly used data mining technique used to discover frequent item sets and association rules in a dataset. For example, by analyzing historical data, an association rule can be discovered, such as "daylily yield is higher when soil moisture is between 60% and 70% and light intensity is greater than 5000 Lux." Association rule mining can help understand the interactions between different environmental factors and their impact on daylily growth.

[0054] A time series analysis algorithm is used to predict trends in daylily growth data, predicting future growth trends based on historical data. Time series analysis is a statistical method used to analyze chronological data sequences and predict future trends. For example, based on daylily yield data from the past few years, time series analysis algorithms such as moving average or exponential smoothing can be used to predict daylily yield trends for the coming year. Time series analysis can help understand daylily growth trends in advance and provide a reference for planting management. The discovered environmental patterns and predicted growth trends are visualized to provide agricultural experts with intuitive data support and assist in developing daylily growth optimization strategies. Visualizations can present complex data intuitively in charts and graphs, making it easier for users to understand and analyze. For example, the association rules between environmental factors and daylily yield can be displayed as a network diagram, and the growth trends of daylilies can be displayed as a line chart. Visualizations can improve the efficiency and effectiveness of data analysis and provide decision support for agricultural experts.

[0055] Based on environmental patterns and growth trends, machine learning algorithms are used to develop a growth optimization model for daylilies. By simulating growth under different environmental conditions, optimal growth conditions are predicted to guide actual planting. Machine learning algorithms can learn patterns from historical data to build predictive models that can be used to predict future outcomes. For example, machine learning algorithms such as support vector machines or neural networks can be used to develop a growth optimization model for daylilies. By inputting different environmental parameters, the model outputs predicted daylily yield or other growth indicators. By continuously adjusting these parameters, the optimal combination of growth conditions can be found, guiding actual planting and improving daylily yield and quality.

[0056] Based on environmental laws and growth trends, a machine learning algorithm was used to establish a daylily growth optimization model. By simulating the growth under different environmental conditions, the optimal growth conditions were predicted to guide actual planting.

[0057] The environmental parameter data and growth status data during the growth process of daylily were obtained, and a mapping relationship dataset between environmental parameters and growth status was established. The decision tree algorithm was used to train the dataset to obtain decision rules between the growth status and environmental parameters of daylily. According to the decision rules, the key environmental parameters affecting the growth of daylily were determined, and the optimal value range of each key parameter was determined. The support vector machine algorithm was used to train the growth prediction model with key environmental parameters as features and growth status as the target. The genetic algorithm was used to solve the optimal growth condition combination with the value range of key environmental parameters as constraints and the output of the growth prediction model as the fitness function. The values ​​of each key environmental parameter under the optimal growth condition combination were used as the target values ​​for the daylily planting environment control. During the actual planting process, the planting environment parameters were dynamically adjusted according to the deviation between the environmental monitoring data and the environmental control target value to make them approach the optimal growth conditions, thereby guiding the optimized planting of daylily.

[0058] Specifically, environmental monitoring data is collected during the growth of the daylily, such as soil temperature, humidity, nitrogen, phosphorus, and potassium content, light intensity, air temperature, humidity, and carbon dioxide concentration. This data is then mapped to the plant height, number of leaves, and number of flower buds, forming a dataset. For example, at a specific moment, the soil temperature is monitored to be 25 degrees Celsius, the soil moisture is 60%, and the light intensity is 800 micromoles per square meter per second. Correspondingly, the plant height, number of leaves, and number of flower buds are 50 centimeters, 15, and 5 at this time. This data set is recorded as a sample in the dataset. Over time, data is continuously collected at different time points, expanding the dataset and providing a data foundation for subsequent analysis. A decision tree algorithm is used to train the dataset. Based on the existing data, the algorithm automatically constructs a tree-like model for predicting the growth status of the daylily. Each node of this tree represents an environmental parameter, each branch represents a range of values ​​for that parameter, and each leaf node represents a growth status. For example, the root node of a decision tree might be soil temperature. Branches below 20°C lead to a leaf node, indicating that low temperatures slow growth. Branches above 30°C lead to another leaf node, indicating that high temperatures may cause wilting. Branches between 20 and 30°C lead to another internal node, perhaps light intensity. Different ranges of light intensity further subdivide growth states. In this way, the decision tree algorithm can reveal the complex relationship between environmental parameters and growth states. By analyzing different decision trees within the trained decision tree model, the key environmental parameters affecting daylily growth can be identified and the optimal range for each parameter determined. For example, the decision tree model might reveal that daylily plant height growth is fastest and the number of flower buds is greatest when soil temperature is between 22 and 28°C and light intensity is between 600 and 1000 micromoles per square meter per second. Therefore, soil temperature and light intensity are the key environmental parameters, and 22 to 28°C and 600 to 1000 micromoles per square meter per second are their respective optimal ranges. Such results can help growers understand which environmental factors are most important for daylily growth and how to manipulate these factors for optimal yields.

[0059] Furthermore, the obtained key environmental parameters are further subjected to the support vector machine algorithm to determine a method for obtaining the optimal value:

[0060] A support vector machine (SVM) algorithm was used to train a model that can predict the growth status of daylilies based on key environmental parameters. SVMs excel at finding the optimal interface between different data categories. Here, different growth states, such as "good growth," "slow growth," and "disease presence," can be considered as distinct categories. The SVM algorithm then finds a model that can accurately distinguish these categories based on key environmental parameters such as soil temperature and light intensity. For example, given the current soil temperature and light intensity, the model can predict the probability that the daylily is currently in the "good growth" state. A genetic algorithm was used to find the optimal combination of growth conditions. Genetic algorithms mimic the process of biological evolution, gradually optimizing the solution through continuous iteration. Here, the value of each key environmental parameter can be considered a "gene," a set of environmental parameter combinations can be considered an "individual," and the output of the growth prediction model—the growth status of the daylily—can be considered the "fitness" of the individual. For example, an "individual" might have a soil temperature of 25 degrees Celsius and a light intensity of 800 micromoles per square meter per second. If this combination enables the growth prediction model to predict optimal growth conditions, then the fitness of this "individual" is high. The genetic algorithm simulates operations such as selection, crossover, and mutation to continuously generate new "individuals," retaining those with high fitness, and ultimately finding the optimal combination of growth conditions. The values ​​of key environmental parameters under this optimal growth condition combination are used as target values ​​for controlling the daylily cultivation environment. For example, if the genetic algorithm ultimately determines that a soil temperature of 26 degrees Celsius, a light intensity of 900 micromoles per square meter per second, and a soil moisture of 65% are the optimal growth condition combination, these values ​​can be used as target values ​​for environmental control. During the actual planting process, sensors monitor environmental parameters in real time, compare the monitored data with the target values, and dynamically adjust the planting environment based on any deviations. For example, if the current soil temperature is monitored at 24 degrees Celsius, lower than the target of 26 degrees Celsius, heating equipment can be used to raise the soil temperature. If the current light intensity is monitored at 700 micromoles per square meter per second, lower than the target of 900 micromoles per square meter per second, artificial lighting can be used to increase the light intensity. In this way, through continuous monitoring and adjustment, the planting environment is always kept close to the optimal growth conditions, guiding relevant personnel to adjust relevant conditions, achieve optimized daylily cultivation, and improve yield and quality.

[0061] Step S104: Comprehensively use machine learning algorithms such as support vector machines and random forests to model and analyze environmental monitoring data and daylily growth data to train a model for predicting daylily growth conditions. This includes:

[0062] The above optimal values ​​are used as the control conditions to guide planting, and environmental monitoring data and daylily growth data are obtained in real time. The data are preprocessed, including data cleaning, feature extraction, and data standardization, to obtain a data set suitable for modeling and analysis.

[0063] Based on the preprocessed dataset, a support vector machine algorithm was used for modeling and training. By adjusting the algorithm parameters and selecting features, a support vector machine model that fits the data well was obtained. Based on the prediction results of the support vector machine model, a random forest algorithm was used for ensemble learning. By constructing multiple decision trees and combining their prediction results, the model's generalization ability and prediction accuracy were improved.

[0064] The trained support vector machine model and random forest model were applied to new environmental monitoring data to predict the growth status of daylily under the environmental conditions, and the prediction results were visualized.

[0065] The model's predictive performance is evaluated by comparing actual daylily growth data with the model's predictions. If the prediction error exceeds a preset threshold, the model parameters are adjusted and retrained. Based on the model's predictions, the system determines whether the current environmental conditions are suitable for daylily growth. If not, it recommends measures to improve the environment, such as adjusting temperature and humidity and supplementing lighting. Environmental monitoring data and daylily growth data are continuously collected, and the model is regularly updated and optimized to adapt to changes in the environment and daylily varieties, maintaining stable model predictive performance.

[0066] Specifically, obtaining environmental monitoring data and daylily growth data is the foundation for building a predictive model. Environmental monitoring data includes factors such as soil temperature, humidity, nitrogen, phosphorus, and potassium content, light intensity, air temperature, humidity, and carbon dioxide concentration. Daylily growth data includes indicators such as plant height, number of leaves, and number of flower buds. Data preprocessing is a key step in ensuring effective model training. Data cleaning can remove outliers, such as discarding temperature data that falls outside the normal range.

[0067] Feature extraction can extract more valuable information from raw data, such as calculating average daily temperature and cumulative sunshine time. Data normalization can eliminate dimensional differences between different features, making the model more likely to converge.

[0068] The related support vector machine (SVM) is a powerful classification and regression algorithm. In daylily growth prediction, key environmental parameters are used as input features, and growth status, including indicators such as plant height, number of leaves, and number of flower buds, is used as the output. By adjusting the kernel function type and penalty parameters, better fitting results can be achieved. For example, using a radial basis function (RBF) kernel may be more suitable than a linear kernel to capture the nonlinear relationship between environmental factors and growth status.

[0069] The output of the support vector machine is used as the input to the random forest algorithm. By constructing multiple decision trees and integrating their prediction results, the random forest algorithm can effectively reduce the risk of overfitting. In the prediction of daylily growth, the number of trees and maximum depth can be set appropriately to balance the complexity and generalization ability of the model. For example, the optimal number of trees can be determined through cross-validation, such as 500 trees with a maximum depth of 10. During the model application phase, new environmental monitoring data can be input into the trained model to predict the growth status of daylilies and determine the key environmental parameters that have a significant impact on daylilies. The prediction results can be visualized through line charts, heat maps, and other methods to intuitively display the expected growth of daylilies under different environmental conditions.

[0070] Model evaluation is crucial for ensuring prediction accuracy. Prediction error can be measured using metrics such as the root mean square error (RMSE) or mean absolute error (MAE). If the error exceeds a preset threshold, such as an RMSE greater than 0.5, the model parameters must be adjusted and retrained. This may involve adding training data, adjusting feature selection, or modifying the model structure. Based on the model's predictions, specific recommendations can be made for daylily cultivation. For example, if the model predicts slow growth at current temperatures, a recommendation could be made to raise the greenhouse temperature by 2-3°C. If insufficient light is predicted to affect bud formation, a recommendation could be made to increase the duration of supplemental lighting, providing an additional 2-3 hours of artificial light daily. Continuous model updating and optimization are key to maintaining prediction accuracy. New environmental monitoring data and daylily growth data can be collected monthly and the model retrained. This allows the model to adapt to factors such as seasonal changes and the introduction of new varieties. Through regular updates, the model can continuously learn new growth patterns and improve its predictive capabilities for daylily growth under diverse environmental conditions.

[0071] Step S105: Based on the prediction model, intelligent prediction and decision optimization of the daylily growth process are realized to provide a basis for scientific planting management.

[0072] Sensors collect environmental monitoring data on the daylily's growing environment, including soil temperature, humidity, nitrogen, phosphorus, and potassium content, light intensity, air temperature, humidity, and carbon dioxide concentration, and transmit this data to a data processing center. For more comprehensive monitoring and analysis, forecasts from meteorological authorities are used to obtain information on future weather conditions, such as temperature and precipitation, and to capture images of the daylily's growth.

[0073] Image recognition technology is used to analyze images taken during the growth of daylily to identify the growth status, leaf color, pest and disease conditions of the daylily. Soil nutrient data, including nitrogen, phosphorus, potassium content, etc., are collected using soil sensors to determine the soil fertility. Environmental monitoring data, meteorological data, image recognition results, soil nutrient data, etc. are input into the above-mentioned daylily growth prediction model, and the growth trend and yield of the daylily are predicted through a machine learning algorithm. The prediction model can use the prediction model of step S104 as an extended prediction model to add relevant input ports and output ports, and retrain.

[0074] Based on the prediction results and combined with planting management experience from the expert knowledge base, rule-based engine technology is used to generate scientific planting decision plans, such as irrigation time and water usage, fertilization time and amount, and pest and disease control measures. These decisions are then pushed to farmers in real time via mobile apps or text messages, guiding them in precise planting management operations, thereby achieving intelligent and optimized daylily production.

[0075] Specifically, weather forecast data is an indispensable source of information for agricultural production. For example, the meteorological department predicts continuous rainfall over the next three days, with cumulative precipitation expected to reach 50 mm and temperatures to remain between 20°C and 28°C. This information can help farmers predict weather conditions over the next few days and prepare countermeasures in advance. For example, based on the rainfall forecast, irrigation can be appropriately reduced to avoid excessive soil moisture that can lead to root diseases. Combined with weather forecast data, agricultural activities can be planned more scientifically, improving agricultural production efficiency and resilience to risks.

[0076] Image recognition technology has broad application prospects in monitoring the growth of daylilies. For example, by using a high-resolution camera to regularly capture images of daylily growth, image processing algorithms can automatically identify growth indicators such as plant height and leaf area. For example, identifying an average plant height of 60 cm and dark green leaves from an image indicates healthy growth. Image recognition can also be used for early detection of pests and diseases. For example, by analyzing the characteristics of spots in leaf images, leaf spot can be identified with an accuracy rate exceeding 85%. Prompt detection of pests and diseases allows for early prevention and control measures, minimizing yield losses.

[0077] Soil nutrients are a key factor influencing the yield and quality of daylilies. Soil sensors can monitor the real-time content of nutrients such as nitrogen, phosphorus, and potassium in the soil. For example, soil nitrogen levels of 1.5 grams per kilogram, phosphorus levels of 0.8 grams per kilogram, and potassium levels of 15 grams per kilogram can reflect soil fertility. Based on this data and the nutrient requirements of daylilies at different growth stages, a precise fertilization plan can be developed. For example, during the bud stage of daylilies, the application of phosphorus and potassium fertilizers should be appropriately increased to promote bud development. Precision fertilization can improve fertilizer utilization, reduce production costs, and minimize environmental pollution.

[0078] By inputting the aforementioned data, including environmental monitoring data, image recognition results, meteorological data, and soil nutrient data, into a pre-trained machine learning model, we can predict the growth trends and yields of daylilies. For example, the model predicts an average daily growth of 2 cm for the next week, with an estimated yield per mu of 1,200 kg. These predictions provide a scientific basis for planting decisions. The accuracy of model predictions directly impacts the effectiveness of decision-making, so continuous optimization of model parameters is necessary to improve prediction accuracy. Based on the growth trend, the growth stage of the daylily can be determined.

[0079] Based on the knowledge base of agricultural experts, rule engine technology can automatically generate planting decision plans based on predictions. For example, if the model predicts high temperatures and low soil moisture for the next week, the rule engine can generate the following decision plan: irrigate once in the morning and evening each day, with each irrigation volume of 10 cubic meters per mu; and spray the leaves with a 0.2% potassium dihydrogen phosphate solution to enhance drought resistance.

[0080] These decision-making plans, developed based on expert experience and scientific principles, are highly targeted and actionable. By delivering these plans to farmers via mobile apps or text messages, precise planting management guidance can be achieved. For example, a farmer might receive a text message reminding them, "High temperatures and little rain are expected over the next three days. Please ensure timely irrigation and proper pest and disease control." Based on these instructions, farmers can promptly implement appropriate management measures. This approach ensures farmers receive the latest planting guidance information, improving the efficiency and accuracy of planting management. This intelligent planting management approach can significantly increase the yield and quality of daylilies, increase farmers' income, and promote agricultural modernization.

[0081] Based on the prediction results and combined with the planting management experience in the expert knowledge base, rule engine technology is used to generate scientific planting decision-making plans, such as irrigation time and water usage, fertilization time and amount, and pest and disease control measures.

[0082] Based on the prediction results, data such as crop growth stages and environmental monitoring parameters are obtained. This data is then combined with planting management experience from an expert knowledge base to construct a planting decision rule library. A fuzzy inference algorithm is used to match the obtained crop growth stages and environmental monitoring parameters with the planting decision rule library to infer the appropriate irrigation time and water volume. Based on the crop growth stage and soil nutrient status, fertilization management experience is obtained from the expert knowledge base to construct a fertilization decision rule library.

[0083] A forward chain inference algorithm uses crop growth stage and soil nutrient status as input, matching them with a fertilization decision rule base to infer the appropriate time and amount of fertilization. Image recognition technology is used to capture images of crop pest and disease symptoms. This information is combined with pest and disease control experience from an expert knowledge base to construct a knowledge base for pest and disease diagnosis and control. A case-based reasoning intelligent diagnosis algorithm is used to match similarity between captured crop pest and disease symptom images and cases in the pest and disease diagnosis and control knowledge base to infer the optimal control measures. Irrigation, fertilization, and pest and disease control decisions are integrated to generate a comprehensive and scientific planting management plan. The results are then fed back to growers to guide precise planting management.

[0084] Specifically, building a rule base for planting decisions is a key component of smart agricultural management. For example, in the case of daylily cultivation, an expert knowledge base might include rules such as "Irrigation should be performed when the average daily temperature exceeds 25°C and the soil moisture content is less than 20%." Fuzzy inference algorithms can handle imprecise inputs, such as categorizing temperature into linguistic variables like "low," "moderate," and "high," making them more aligned with real-world planting experience. Irrigation decisions require consideration of multiple factors. For example, if sensors detect a soil moisture content of 18% and the weather forecast predicts no rainfall for the next three days, the fuzzy inference system might conclude, "Immediately implement moderate-intensity irrigation." This approach is more accurate than traditional fixed-cycle irrigation and effectively conserves water resources. Fertilization decisions are also based on multi-dimensional data. If soil testing indicates low nitrogen content and the daylily is in its vegetative growth phase, forward chain reasoning might deduce the decision to "Apply a compound fertilizer with a 40% nitrogen content at a rate of 20 kg per mu." This precision fertilization not only improves fertilizer utilization but also reduces environmental pollution. Pest and disease control is a crucial component of planting management. Using image recognition technology, the system might identify symptoms of brown spot disease on daylily leaves. A case-based reasoning intelligent diagnostic algorithm searches the knowledge base for similar cases. If it finds a case with a 90% similarity, the control solution is "spraying 25% azoxystrobin suspension, diluted 500 times." The system then provides this solution as a recommendation to the grower. Generating a comprehensive decision requires balancing various management measures. For example, if the system recommends both irrigation and pest control, it considers that immediate irrigation after spraying may reduce the efficacy of the pesticide and may therefore recommend "perform pest control first, followed by irrigation 24 hours later." This comprehensive consideration maximizes the effectiveness of each management measure.

[0085] The advantages of intelligent decision-making systems lie in their rapid response and continuous optimization capabilities. For example, if the system detects that daylily growth is slower than expected, it immediately analyzes possible causes, such as insufficient nutrients or light, and recommends appropriate adjustments. As data accumulates, the system's prediction accuracy will continue to improve, and its decisions will become increasingly precise. However, intelligent systems do not completely replace human decision-making; rather, they serve as auxiliary tools. Farmer experience remains crucial, and the system's decision-making solutions should be integrated with farmers' actual practices. For example, the system might recommend irrigation during specific time periods, but the specific irrigation amount still needs to be fine-tuned by farmers based on their actual conditions. This intelligent, human-machine integration management model can significantly improve daylily cultivation efficiency and yield. For example, after implementing this system in a pilot field, water use decreased by 20%, fertilizer use by 15%, and yield increased by 10%. This not only brings economic benefits but also promotes sustainable agricultural development. The application of intelligent decision-making systems is gradually transforming traditional agriculture, driving it towards precision and intelligent farming.

[0086] Step S106, develop a visualization system for the growth supervision of daylilies, which uses data visualization technology to intuitively display the real-time dynamics of the growth environment and growth status of daylilies, and present the prediction results in the form of charts or alarms, providing farmers with intuitive and easy-to-use data query and decision support functions.

[0087] By integrating the prediction results and early warning information into the visualization system in the form of charts or alarms, farmers can intuitively view the changing trends of the daylily growth environment in the future through the system interface, and take corresponding measures in a timely manner according to the early warning prompts, such as adjusting irrigation, fertilization, pest and disease control, etc., to ensure that the daylily grows under optimal conditions.

[0088] Provide farmers with convenient data query capabilities. Through the visualization system's interactive interface, farmers can flexibly query historical daylily growth data by time period, key indicators, and other dimensions, understand growth conditions at different times, conduct horizontal or vertical comparative analysis, and grasp the patterns and characteristics of daylily growth. Step 7: Based on the real-time data, forecast information, and historical data query capabilities provided by the daylily growth monitoring visualization system, farmers are provided with scientific decision-making support. Based on the system's data analysis results, farmers can optimize daylily planting and management plans, such as optimizing planting times, rationally arranging irrigation and fertilization, and promptly preventing and controlling pests and diseases, thereby improving daylily yield and quality and achieving precise and intelligent daylily production management.

[0089] Specifically, the visual interface design is a crucial component of the system. For example, a line chart can be used to display temperature trends, a heat map can be used to depict soil moisture distribution, and a dashboard can be used to visually display light intensity. This design allows farmers to quickly understand the growing environment conditions for their daylilies. Furthermore, color-coded alerts can be set. For example, when the temperature exceeds 35°C, the relevant data will be displayed in red, alerting farmers to the effects of high temperatures. The development of predictive models is crucial for proactively responding to adverse growing conditions. For example, a support vector machine model trained using historical data can predict temperature fluctuations over the next 24 hours. If the forecast indicates that the temperature may drop below 5°C the next day, the system will automatically issue a low-temperature warning, reminding farmers to prepare for the cold. Integrating prediction results and warning information into the visualization system can help farmers better plan their planting management. For example, if the system predicts continued high temperatures over the next week, farmers can adjust their irrigation plans in advance, increasing irrigation frequency or water volume to mitigate potential drought stress.

[0090] The historical data query function provides farmers with a tool for in-depth analysis. For example, farmers can compare growth conditions during the same period across different years, such as viewing the average temperature and light intensity for May last year and this year, to understand the impact of climate change on daylily growth. This horizontal comparison helps farmers summarize experience and optimize planting strategies. Based on the data and analysis results provided by the system, farmers can develop more scientific planting management plans. For example, historical data analysis revealed that daylilies grow best in temperatures of 20-25°C and relative humidity of 60-70%. Based on this finding, farmers can adjust greenhouse environmental control parameters to create the most suitable growing environment. Alternatively, if the system predicts continuous rainy weather in the coming days, farmers can adjust fertilization plans in advance to prevent nutrient loss and improve fertilizer utilization efficiency. Through these specific implementation methods, the daylily growth monitoring visualization system can comprehensively improve the accuracy and scientific nature of planting management, effectively increase daylily yield and quality, and realize an intelligent and digital modern agricultural production model.

[0091] like Figure 4 As shown, the present invention also provides a daylily field management optimization system based on big data, which mainly includes:

[0092] The environmental data acquisition module is used to monitor the temperature, humidity, light intensity, and soil moisture of the daylily's growing environment using multiple types of sensors to obtain comprehensive environmental data.

[0093] The data fusion processing module is used to integrate and collaboratively process multi-source heterogeneous monitoring data from different sensors through sensor fusion technology to achieve effective data integration and obtain accurate environmental monitoring results;

[0094] The data management and analysis module is used to build a data management system on the cloud computing platform, perform pre-processing operations such as classification, storage, cleaning and filtering on massive monitoring data, and use big data analysis technology to mine environmental patterns and growth trends from the data, providing data support for the optimization of daylily growth;

[0095] The growth model training module is used to comprehensively use machine learning algorithms such as support vector machines and random forests to model and analyze environmental monitoring data and daylily growth data, and train a model to predict the growth status of daylilies;

[0096] An intelligent prediction and decision-making module is used to realize intelligent prediction and decision optimization of the daylily growth process based on the prediction model, providing a basis for scientific planting management;

[0097] The visualization supervision module is used to develop a visualization system for the growth supervision of daylilies. It uses data visualization technology to intuitively display the real-time dynamics of the daylily's growth environment and growth status, and presents the prediction results in the form of charts or alarms, providing farmers with intuitive and easy-to-use data query and decision-making support functions.

[0098] The above are merely preferred embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A method for optimizing daylily field management based on big data, characterized in that: include: Acquire environmental monitoring data of a daylily field, mine the environmental monitoring data through big data analysis to obtain control data, and guide environmental control of the daylily according to the control data; Predicting the environmental monitoring data using a machine learning algorithm to obtain expected growth conditions of the daylily; Acquiring meteorological data and image data of the daylily, obtaining disease data and growth status based on the image data, predicting the growth stage and yield of the daylily based on the meteorological data, environmental monitoring data, disease data, and growth status, and using a rule engine to identify and optimize based on the growth stage, yield, environmental monitoring data, and disease data to obtain a control plan for further regulating field management of the daylily; Based on the prediction results and combined with the planting management experience in the expert knowledge base, rule engine technology is used to generate scientific planting decision plans, including irrigation time and water amount, fertilization time and amount, and pest and disease control measures; Push decision-making plans in real time via mobile apps or text messages, thus realizing intelligent and optimized daylily production processes; High-resolution cameras are used to regularly capture images of the daylily's growth, and image processing algorithms are used to automatically identify the plant height and leaf area. Image recognition is also used for early detection of pests and diseases. Soil sensors monitor the nitrogen, phosphorus, and potassium nutrient content in the soil in real time. Based on this data and the nutrient requirements of daylilies at different growth stages, precise fertilization plans are developed. Various data, including environmental monitoring data, image recognition results, meteorological data, and soil nutrient data, are input into a pre-trained machine learning model to predict the growth trend and yield of daylilies. Rule engine technology automatically generates planting decision plans based on the prediction results based on the knowledge base of agricultural experts; These decision-making plans are based on expert experience and scientific principles and are pushed to farmers via mobile apps or text messages, providing precise planting management guidance. Based on the prediction results and combined with the planting management experience in the expert knowledge base, rule engine technology is used to generate scientific planting decision plans; Based on the prediction results, the crop growth stage and environmental monitoring parameter data are obtained. Combined with the planting management experience in the expert knowledge base, a planting decision rule base is constructed. Using a fuzzy inference algorithm, the obtained crop growth stage and environmental monitoring parameters are used as input, matched with the planting decision rule base, and the appropriate irrigation time and water volume are inferred. Based on the crop growth stage and soil nutrient status, fertilization management experience is obtained from the expert knowledge base to construct a fertilization decision rule base. A forward chain reasoning algorithm is used to take the crop growth stage and soil nutrient status as input, match them with the fertilization decision rule library, and infer the appropriate fertilization time and amount. The image of crop disease and pest symptoms is obtained through image recognition technology. Combined with the pest and disease prevention experience in the expert knowledge base, a pest and disease diagnosis and prevention knowledge base is constructed. An intelligent diagnosis algorithm based on case reasoning is used to match the obtained crop disease and pest symptom images with the cases in the pest and disease diagnosis and prevention knowledge base for similarity, and infer the optimal prevention and control measures. Irrigation decisions, fertilization decisions and pest and disease prevention decisions are integrated to generate a scientific and complete planting management decision plan, and the decision results are fed back to growers to guide precise planting management. The construction of the planting decision rule library is a key link in intelligent agricultural management.

2. The method according to claim 1, characterized in that The environmental monitoring data include soil temperature, humidity, nitrogen, phosphorus and potassium content, light intensity, air temperature, humidity and carbon dioxide concentration.

3. The method according to claim 1, characterized in that Before mining the environmental monitoring data, the following steps are also included: The environmental monitoring data is preprocessed, wherein the preprocessing includes noise removal and data cleaning and filtering.

4. The method according to claim 1, wherein Before mining the environmental monitoring data, the following steps are also included: The environmental monitoring data is distributedly stored in a database, wherein the environmental monitoring data is classified by a machine learning method, and the environmental monitoring data is stored according to the classification results.

5. The method according to claim 1, wherein The process of mining the environmental monitoring data includes: Obtaining the growth status corresponding to the environmental monitoring data; training the decision tree model using the environmental monitoring data and the growth status to obtain a trained decision tree model, and performing analysis based on the trained decision tree model to obtain key environmental parameters; The machine learning model is trained according to the key environmental parameters and the corresponding growth status, the key environmental parameters are iteratively adjusted through the optimization algorithm, and the key environmental parameters in the iterative adjustment are predicted by the trained machine learning model to obtain the growth status score as the fitness, wherein the key environmental parameters are iteratively adjusted according to the fitness until the stopping condition is reached, and the optimal growth condition combination of the key environmental parameters, i.e., the control data, is obtained.

6. The method according to claim 1, characterized in that The process of predicting the environmental monitoring data includes: The environmental monitoring data is predicted by a support vector machine to obtain the growth data of the daylily, wherein the growth data of the daylily includes plant height, number of leaves, and number of flower buds. The growth data of the daylily is predicted by a random forest model to obtain the expected growth of the daylily.

7. The method according to claim 1, characterized in that The process of obtaining disease data and growth status includes: The image data is identified by an image recognition method to obtain disease data and growth status.

8. The method according to claim 6, characterized in that The growth stages and yield obtaining process of daylily include: The meteorological data, environmental monitoring data, disease data and growth conditions are processed by a machine learning model to obtain the growth stage and yield of the daylily.

9. The method according to claim 1, characterized in that After obtaining the control plan, the method also includes visually displaying the environmental monitoring data, control data, expected growth conditions of the daylily, and the control plan through a visual interface.

10. A daylily field management optimization system based on big data, characterized in that: Used to perform the method according to any one of claims 1 to 9.

Images