Crop planting germination model prediction method based on Internet of Things

Abstract

The invention relates to the technical field of agricultural big data processing, in particular to a crop planting germination model prediction method based on the Internet of Things. The farmland environment and the crop growth condition are monitored in real time, the collected data are transmitted to the big data analysis platform, and advanced data analysis technologies such as a machine learning algorithm are utilized to deeply process and analyze the data, so that the crop yield and the growth condition are predicted more accurately. According to establishment of the crop growth model, the crop growth model is established through a machine learning algorithm such as a random forest and a support vector machine, the crop growth process can be simulated more truly, and the prediction accuracy is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural big data processing, and in particular to a crop planting germination model prediction method based on the Internet of Things. Background Art

[0002] Traditional methods of planting crops in farmland mainly rely on farmers' experience and intuition. It is difficult to accurately control environmental factors and agricultural management measures during the planting process, resulting in waste of resources, unstable yields and uneven product quality. With the development and widespread popularization of Internet of Things technology, the use of intelligent perception, recognition and communication technology can closely connect farmland environmental parameters with the Internet to achieve real-time data collection and transmission. In response to the needs of agricultural informatization and smart agriculture, through big data analysis and algorithm modeling, the germination time and growth status of crops in the planting process can be accurately predicted to provide farmers with scientific agricultural management decision support. It is an important achievement of modern agricultural technology innovation and has far-reaching significance for promoting the process of agricultural modernization. Applying the Internet of Things to farmland crop planting and realizing intelligent management has become a research hotspot in the agricultural field.

[0003] 1. Development of Internet of Things Technology As an important part of the new generation of information technology, the Internet of Things (IoT) combines various information sensing devices with the Internet through communication sensing technologies such as intelligent perception, recognition technology and ubiquitous computing to form a huge network. It connects everything and makes it possible for comprehensive information interaction and seamless integration between people and things, and between things. With the continuous maturity and popularization of IoT technology, its application in various fields is becoming more and more extensive, including the agricultural field.

[0004] 2. Demand for agricultural informatization and smart agriculture With the growth of global population and the increasing prominence of food safety issues, traditional agricultural production models can no longer meet the needs of modern society. The concepts of agricultural informatization and smart agriculture have emerged, aiming to improve agricultural production efficiency, optimize resource allocation, improve agricultural product quality, and promote sustainable agricultural development through information technology. As an important support for agricultural informatization, Internet of Things technology provides strong support for the construction of smart agriculture.

[0005] 3. Importance of crop germination prediction model Crop planting germination is the first step in agricultural production and one of the key factors that determine crop yield and quality. Traditionally, farmers often rely on experience and intuition to judge the germination time and growth status of crops, but this method has great uncertainty and error. The crop planting germination prediction model based on the Internet of Things can accurately predict the germination time and growth status of crops by real-time monitoring of farmland environmental parameters, combined with the biological characteristics of crop growth and historical data. This helps farmers understand the growth of crops in advance, arrange agricultural production activities scientifically and rationally, and improve agricultural production efficiency and output. Summary of the invention

[0006] The present invention provides a crop planting germination prediction model based on the Internet of Things, which realizes real-time monitoring and accurate data collection of farmland environmental parameters by integrating wireless sensor networks, data analysis and algorithm models. This innovative technology not only provides farmers with real-time insights into the growth status of crops, but also can predict the germination time, growth rate and trend of crops with high accuracy by constructing a complex growth prediction model. This not only helps farmers to scientifically and rationally arrange agricultural production activities according to the prediction results, optimize resource allocation, and reduce resource waste, but also enhances the adaptability of agricultural production to climate and environmental changes, enabling farmers to respond to potential growth obstacles and risks in advance. In addition, the introduction of Internet of Things technology has also promoted the intelligent and automated process of agricultural production. Through remote monitoring and intelligent control equipment, farmers can manage farmland more conveniently and improve agricultural production efficiency and quality. In summary, the crop planting germination prediction model based on the Internet of Things is an important breakthrough in modern agricultural technology, and provides strong support for the realization of sustainable agricultural development and intelligent management.

[0007] The present invention solves the deficiencies of the prior art by adopting the following technical solutions: A crop planting germination model prediction method based on the Internet of Things includes the following steps: S1. Data Collection Collect data sources and collect observation data of germinating organisms, such as temperature, humidity, air pressure, wind speed, wind direction, precipitation, soil type, planting method, and seed variety. The data sources are stored in the Hadoop platform. S2. Data cleaning Check the missing values ​​in the data and fill them by filling in the mean value; identify and delete outliers, and standardize or normalize the same observation data to eliminate the dimensional differences between different features; S3. Feature collection: Select some features related to germination prediction in the observed data and convert them into numerical features through label encoding method; S4. Dataset Division The cleaned and preprocessed dataset is divided into training set, validation set and test set. The training set is used to train the model, the validation set is used to adjust the model parameters and select hyperparameters for tuning, and the test set is used to evaluate the performance of the model. S5. Build a random forest model Initialize the model using the machine learning library scikit-learn to initialize the random forest model; set the numerical features as hyperparameters; S6. Model training Training data and building decision trees, each decision tree is trained on a random subset of the training data; S7. Model evaluation and tuning Use the validation set to evaluate the performance of the model and use grid search, random search, or Bayesian optimization to tune the model's hyperparameters. S8. Prediction Prepare new crop planting data, go through the same S2 data cleaning and S3 feature collection steps as the training data, and use the trained random forest model to make predictions on the new data.

[0008] Compared with existing methods, the innovation of the method described in this paper lies in that it can take planting environment factors into account more comprehensively, solving the problem of lack of comprehensive consideration of planting environment factors in the decision-making process, thereby improving the accuracy and quality of decision-making: The present invention not only monitors the farmland environment and crop growth conditions in real time, but also transmits the collected data to a big data analysis platform, and uses advanced data analysis techniques, such as machine learning algorithms, to deeply process and analyze the data, thereby more accurately predicting crop yields and growth conditions. Establishment of crop growth model: The present invention establishes a crop growth model through machine learning algorithms, such as random forests and support vector machines, which can more realistically simulate the crop growth process and improve the accuracy of prediction.

[0009] Based on the prediction results, the present invention combines farmland management and crop growth requirements to provide farmers with targeted planting management suggestions, such as irrigation, fertilization, pest control, etc., thereby realizing personalized farmland management. The present invention reduces the excessive use of pesticides, fertilizers and other chemicals, protects the ecological environment, and realizes the sustainable development of farmland through the rational use of resources.

[0010] The present invention reduces agricultural production costs and improves farmers' economic benefits through scientific and reasonable planting management. Improve farmland management efficiency: The present invention uses Internet of Things technology and big data analysis to achieve real-time monitoring and management of farmland environment and crop growth conditions, greatly improving farmland management efficiency. The present invention combines modern information technology with traditional agriculture to promote the development of agriculture towards modernization and intelligence. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Picture 1 It is the prediction flow chart of the present invention. DETAILED DESCRIPTION

[0012] A crop planting germination model prediction method based on the Internet of Things includes the following steps: S1. Data Collection Collect data sources and collect observation data of germinating organisms, such as temperature, humidity, air pressure, wind speed, wind direction, precipitation, soil type, planting method, and seed variety. The data sources are stored in the Hadoop platform. S2. Data cleaning Check the missing values ​​in the data and fill them by filling in the mean value; identify and delete outliers, and standardize or normalize the same observation data to eliminate the dimensional differences between different features; S3. Feature collection Select some features related to germination prediction in the observed data and convert them into numerical features through label encoding method; S4. Dataset Division The cleaned and preprocessed dataset is divided into training set, validation set and test set. The training set is used to train the model, the validation set is used to adjust the model parameters and select hyperparameters for tuning, and the test set is used to evaluate the performance of the model. S5. Build a random forest model Initialize the model using the machine learning library scikit-learn to initialize the random forest model; set the numerical features as hyperparameters; S6. Model training Training data and building decision trees, each decision tree is trained on a random subset of the training data; S7. Model evaluation and tuning Use the validation set to evaluate the performance of the model and use grid search, random search, or Bayesian optimization to tune the model's hyperparameters. S8. Prediction Prepare new crop planting data, go through the same S2 data cleaning and S3 feature collection steps as the training data, and use the trained random forest model to make predictions on the new data.

[0013] In the S1, some features related to germination prediction in the observed data are regularly replaced, and converted into numerical features through a label encoding method.

[0014] In the S1, observation data of germination organisms are collected periodically.

[0015] Interpretation and Application of Results - Understand how the model makes predictions and explain what the predictions mean. Apply the predictions to real-world scenarios, such as guiding farmers to adjust planting strategies, optimize seed selection, etc. Deploy the trained model to a production environment so that predictions can be made in real-world applications. Monitor the performance of the model regularly and update or retrain the model as needed to maintain its accuracy.

[0016] In order to make the above-mentioned objects, features and advantages of the present invention more understandable, the present invention is further described in detail below through a detailed technical scheme and specific embodiments of corn seed germination.

[0017] Detailed description of technical innovation points: Ensemble learning algorithm: As an ensemble learning algorithm, random forest can effectively process high-dimensional data and complex data relationships, and its robustness is better than that of a single decision tree model. Feature importance analysis: Through the feature importance score of random forest, we can have a deeper understanding of which environmental factors have a greater impact on crop yields and provide data support for subsequent decision-making. Hyperparameter optimization: Using grid search for hyperparameter tuning, we automatically find the optimal model parameters and improve the accuracy and efficiency of model prediction. Feature scaling: Through feature scaling, the convergence speed and prediction performance of the model are optimized, especially when the feature magnitudes vary greatly.

[0018] The detailed technical solution steps of the method for collecting data based on IoT devices and analyzing it through models are as follows: 1. Data Preparation and Preprocessing 1.1 Loading data First, load the collected CSV file.

[0019] Python import pandas as pd # Load data Make sure the dataset format is correct and the file path is correct. The collected data are: Temperature (daily average, minimum, maximum) humidity Soil Type pH rainfall Seed Varieties Seed treatment method (such as soaking, disinfection, etc.) data = pd.read_csv('weather_data.csv') 1.2 Data Cleansing Check whether there are missing values, outliers, or duplicates in the data set and delete them. Cleaning data is a prerequisite for ensuring model quality and ensuring that there are no missing values ​​that affect model training.

[0020] Python # Handle missing values ​​(e.g., fill with mean, median, or mode) data.fillna(data.mean(), inplace=True) # Handle outliers (according to specific circumstances) # ... # Remove duplicates (if any) data.drop_duplicates(inplace=True) 1.3 Feature selection and coding Select the features that have an impact on corn germination prediction and encode the categorical features. Select the features related to the target variable based on the relevant crop knowledge of agricultural species to reduce the interference of irrelevant features.

[0021] Python # Select feature columns # The following features were selected through professional knowledge features = ['temperature_mean', 'temperature_min', 'temperature_max', 'humidity', 'soil_type', 'pH', 'rainfall', 'seedVariety', 'seed Treatment'] target = 'emergence_rate' # Target variable: germination rate # One-hot encode categorical features (if any) # 'soil_type', 'seed Variety', 'seed Treatment' are categorical features label_encoders = {} for feature in ['soil_type', 'seed Variety', 'seed Treatment']: le = LabelEncoder() data[feature] = le.fit_transform(data[feature]) label_encoders[feature] = le 1.4 Feature Scaling Standardize the features to eliminate the impact of different dimensions and improve the efficiency and accuracy of model training. Feature scaling is necessary for some algorithms (such as SVM and KNN). However, it is usually not necessary for random forests because random forests are tree-based algorithms and are not sensitive to the scale of features. Scaling is possible.

[0022] Python from sklearn.preprocessing import StandardScaler scaler = StandardScaler() data[features] = scaler.fit_transform(data[features]) 2. Divide the dataset The dataset is divided into training and testing sets so that the generalization ability of the model can be evaluated on an independent test set.

[0023] Python from sklearn.model_selection import train_test_split # Divide the dataset X_train, X_test, y_train, y_test = train_test_split(data[features], data[target], test_size=0.2, random_state=42) 3. Training the Random Forest Model 3.1 Initialize the model Select regression RandomForestRegressor. Initialize the random forest model and set the random seed to ensure the reproducibility of the results.

[0024] Python from sklearn.ensemble import RandomForestClassifier, RandomForestRegressor # right reg = RandomForestRegressor(n_estimators=100, # number of trees max_depth=None, # Maximum depth of the tree min_samples_split=2, # Minimum number of samples required for internal node re-division random_state=42) # Random seed 3.2 Training Model Use the training data to fit the model and learn the relationship between features and target variables. Here are the specific steps: a. Initialize Model: Create an instance of the selected machine learning algorithm.

[0025] b. Fit the model: Call the model’s training method (usually called fit or train), passing in the feature matrix (X) and the target variable (y).

[0026] Python # Regression Problem reg.fit(X_train, y_train) c. Learning the relationship between features and target variables: During the fitting process, the model learns the relationship between features and target variables. This process includes the following sub-steps: i. Parameter estimation: The model estimates the model parameters through optimization algorithms (gradient descent, least squares method) to minimize the prediction error.

[0027] ii. Feature selection: Feature selection is performed during the training process to determine which features are most important for predicting the target variable.

[0028] iii. Model structure determination: For complex models such as neural networks, the training process also includes determining the network structure, such as the number of layers, number of neurons, etc. 4. Evaluate the model Use the test set data to evaluate the model performance and calculate the mean square error to quantify the accuracy of the prediction. The following are the specific steps: a. Calculate the square of the error: Square the error of each sample to eliminate the offsetting effect between positive and negative errors.

[0029] b. Calculate the mean square error: sum the squared errors of all samples and divide by the number of samples to get the mean square error. The formula for MSE is: Where n is the number of samples in the test set, yi It is i The true target variable value of samples, hatyi The model is i The predicted value of a sample.

[0030] c. Interpretation of MSE results The smaller the MSE value, the closer the model's prediction result is to the true value, that is, the higher the model's prediction accuracy is.

[0031] The numerical value of MSE can intuitively reflect the average level of model prediction error.

[0032] python from sklearn.metrics import accuracy_score, mean_squared_error, classification_report, confusion_matrix # Regression Problem y_pred = reg.predict(X_test) mse = mean_squared_error(y_test, y_pred) print(f'Mean Squared Error: {mse}') 5. Hyperparameter Tuning Use grid search, random search, or Bayesian optimization to find the best model hyperparameters to optimize the model and find the best parameter combination to improve model performance.

[0033] Python from sklearn.model_selection import GridSearchCV, RandomizedSearchCV # Define the hyperparameter grid to search param_grid = { 'n_estimators': [100, 200, 300], 'max_depth': [None, 5, 10, 15], 'min_samples_split': [2, 5, 10], # ... other hyperparameters } 6. Predicting New Data Use the trained model to predict new crop planting data. And use the optimal model to predict and get the crop yield prediction value.

[0034] Python #data new_data # new_data = ... # new_data_predictions = reg.predict(new_data) # For regression problems 7. Result interpretation and visualization Explain the model prediction results and visualize the feature importance to help understand which features have the greatest impact on the prediction results.

[0035] Python # For feature importance importances = clf.feature_importances_ std = np.std([tree.feature_importances_ for tree in clf.estimators_], axis=0) indices = np.argsort(importances)[::-1] # Print the feature ranking print("Feature ranking:") for f in range(X.shape[1]): print("%d. feature %s (%f)" % (f + 1, X.columns[indices[f]],importances[indices[f]])) 10. The conclusion obtained from the model calculation is Daily average temperature: The daily average temperature for corn seed germination is between 20°C and 25°C. Too low a temperature will delay germination or even prevent the seeds from germinating; too high a temperature may cause heat damage to the seeds and reduce the germination rate. Minimum temperature: Avoid too low a temperature at night, usually not less than 13°C, to prevent chilling damage. Maximum temperature: Avoid too high a temperature during the day, which may inhibit germination if it exceeds 25°C.

[0036] Soil moisture is crucial to germination. Appropriate soil moisture (usually 55%-70% of field water holding capacity) is conducive to seed water absorption and swelling, promoting germination. Too low humidity will lead to insufficient water absorption and difficulty in germination; too high humidity may cause seed rot.

[0037] The effects of different soil types on corn seed germination are mainly reflected in soil structure, water retention and air permeability. Loose, air-permeable and moderately water-retaining clay soil is conducive to seed germination.

[0038] Corn has a strong adaptability to soil pH, but the most suitable pH range is between 6.3 and 6.9. Too acidic or alkaline soils may affect the germination rate of seeds.

[0039] Adequate rainfall or irrigation can help maintain soil moisture and promote seed germination, but too much rain may cause waterlogging in the soil, affecting seed respiration and germination.

[0040] Different varieties of corn seeds have different adaptability to the environment. Some varieties are more resistant to adverse conditions such as low temperature, drought or salinity, so the germination rate will also be different.

[0041] Appropriate seed treatment (such as soaking, disinfection, etc.) can improve the germination rate and stress resistance of seeds. For example, warm water soaking can promote seed water absorption, and disinfectants can kill pathogens on the surface of seeds.

[0042] Technical points of the present invention: Real-time monitoring of farmland environment and crop growth conditions: By deploying a variety of sensors in the farmland, such as soil moisture sensors, temperature sensors, light sensors, etc., relevant data on the farmland environment and crop growth conditions can be collected in real time.

[0043] Data transmission and processing: The monitored data is transmitted to the big data analysis platform through the Internet of Things. The platform pre-processes the data, including data cleaning and denoising, to ensure data quality.

[0044] Establish crop growth model: Use machine learning algorithms, such as random forest and support vector machine, to analyze the processed data, establish crop growth model and simulate the crop growth process. The following is an overview of the specific implementation process: 1. Model Construction Select the algorithm: Choose the appropriate machine learning algorithm based on the nature of the problem and the characteristics of the data. For complex nonlinear relationships, random forests are often used because they can handle high-dimensional data and automatically select features; support vector machines (SVMs) perform well in handling small samples, high-dimensional data, and nonlinear classification.

[0045] Divide the dataset: Divide the dataset into training set, validation set and test set, usually in a ratio of 70%:15%:15% or similar.

[0046] Train model: Use the training set data to train the selected machine learning model and adjust the model parameters to optimize performance.

[0047] Feature importance assessment: In models such as random forests, assessing the importance of each feature to the model’s prediction results helps understand which factors have the greatest impact on crop growth.

[0048] 2. Model validation and optimization Cross-validation: K-fold cross-validation and other methods are used to evaluate the stability and generalization ability of the model.

[0049] Hyperparameter tuning: Use methods such as grid search, random search, or Bayesian optimization to adjust model hyperparameters to further improve model performance.

[0050] Model comparison: Compare the prediction effects of different machine learning algorithms (such as random forest and support vector machine) and select the optimal model.

[0051] 3. Model application and simulation Crop growth simulation: Use the trained model to input specific environmental conditions and agricultural management strategies to simulate the growth process of crops and predict crop growth trends, yields and quality.

[0052] Decision support: Based on simulation results, scientific advice is provided to farmers on irrigation, fertilization, pest and disease control and other farming activities to optimize crop growth conditions and improve yield and quality.

[0053] Continuous optimization: As new data is collected, the model is updated regularly to maintain its predictive accuracy and timeliness.

[0054] 4. Feedback and Evaluation Field validation: Compare the model prediction results with field observation data to evaluate the accuracy and practicality of the model.

[0055] User feedback: Collect feedback from farmers and agricultural experts to understand the problems and improvement directions of the model in practical applications.

[0056] Continuous improvement: Based on evaluation results and user feedback, we continuously optimize model algorithms and data sets to improve the adaptability and accuracy of the model.

[0057] Predict crop yield and growth conditions: Based on the established crop growth model, input real-time monitoring data, predict crop yield and growth conditions, and output prediction results.

[0058] Provide planting management suggestions: Based on the prediction results, combined with farmland management and crop growth needs, provide farmers with scientific and reasonable planting management suggestions, such as irrigation, fertilization, pest and disease control, etc.

[0059] Beneficial effects of the invention: Improve farmland yields: Through real-time monitoring and management, ensure that the crop growth environment is suitable and increase crop yields; the present invention uses big data analysis and machine learning algorithms to accurately predict crop growth conditions and provide farmers with targeted planting management suggestions, thereby increasing farmland yields.

[0060] Improve the sustainability of farmland: Rationally utilize resources, reduce excessive use of chemicals such as pesticides and fertilizers, and protect the ecological environment; the present invention realizes precise management of resources, reduces excessive use of chemicals, and reduces negative impacts on the ecological environment through real-time monitoring of the farmland environment and crop growth conditions.

[0061] Reducing agricultural production costs: reducing agricultural production costs through scientific and reasonable planting management; the present invention provides farmers with targeted planting management suggestions to help farmers optimize the agricultural production process and reduce production costs.

[0062] Improve farmers' income: Increase crop yields and increase farmers' income; the present invention helps farmers increase their income and improve their living standards by increasing farmland yields.

[0063] Improve farmland management efficiency: The present invention utilizes Internet of Things technology and big data analysis to achieve real-time monitoring and management of farmland environment and crop growth conditions, thereby improving farmland management efficiency.

[0064] Promoting the modernization of agriculture: This invention combines modern information technology with traditional agriculture, promotes the development of agriculture towards modernization and intelligence, and improves the competitiveness of my country's agriculture.

[0065] In summary, the present invention has significant innovation and practicality in the field of farmland management and crop planting prediction, which helps to improve farmland yield and sustainability, reduce agricultural production costs, increase farmers' income, and promote agricultural modernization.

[0066] This patent solution can achieve Precision environmental monitoring: Real-time monitoring of key environmental parameters for crop growth, such as temperature, humidity, light intensity, soil moisture and nutrients, through IoT sensor systems.

[0067] Data collection and processing: Integrate data acquisition equipment to collect large amounts of environmental data, and use patented algorithms or software to perform data cleaning, feature extraction, and analysis to extract useful information.

[0068] Germination prediction model: Use machine learning algorithms to build a prediction model to predict the germination probability of crop seeds based on historical data and real-time monitoring data.

[0069] Decision support system: Provides planting decision support based on prediction results, such as optimal sowing time, planting site selection, optimal resource allocation, etc.

[0070] Automated control system: Integrated automated control system that automatically adjusts environmental conditions such as temperature and humidity control, irrigation system, etc. according to the predictive model to optimize the crop growth environment.

[0071] User Interface: Provide a user-friendly interface to enable farmers to easily access and understand the forecast results and take appropriate actions.

[0072] Remote monitoring and maintenance: Remotely monitor the system operation status through the cloud platform or mobile application, and perform remote maintenance and upgrades.

[0073] Possible future application scenarios Precision agriculture: Through real-time monitoring and prediction, precise control of crop planting can be achieved to improve crop yield and quality.

[0074] Climate change adaptation: predict the impact of climate change on crop growth, develop adaptive planting strategies, and reduce the impact of climate change on agriculture.

[0075] Resource optimization: Based on the prediction results, rationally allocate resources such as water, fertilizers and pesticides to reduce waste.

[0076] Pest and disease prevention: predict environmental conditions that may lead to pests and diseases, take preventive measures in advance, and reduce losses.

[0077] Smart agriculture: Integrate more IoT devices and machine learning technologies to achieve automation and intelligence in agricultural production.

[0078] Agricultural research: Provide data support for agricultural scientists to study the relationship between crop growth and environmental factors, and promote the development of agricultural science.

[0079] Agricultural product quality control: Through predictive models, we can understand which environmental factors affect the quality of agricultural products, so as to control these factors and improve the quality of agricultural products.

[0080] Glossary: Internet of Things - refers to a network that uses information sensing devices to exchange and communicate information on any object in accordance with agreed protocols to achieve intelligent identification, positioning, tracking, monitoring and management.

[0081] Smart farmland refers to modern farmland that uses modern information technology means such as the Internet of Things, cloud computing, and big data to achieve real-time monitoring of farmland environmental parameters, data analysis, and decision support, thereby improving farmland management levels and crop yields.

[0082] Crop planting forecast refers to predicting crop growth trends and yields through real-time monitoring and data analysis of farmland environmental parameters and crop growth status, providing a scientific basis for farmland management.

Claims

1. A crop planting germination model prediction method based on the Internet of Things, characterized in that The steps include: S1. Data Collection Collect data sources and collect observation data of germinating organisms, such as temperature, humidity, air pressure, wind speed, wind direction, precipitation, soil type, planting method, and seed variety. The data sources are stored in the Hadoop platform. S2. Data cleaning: Check the missing values ​​in the data and fill them in by filling the average value; identify and delete outliers, and standardize or normalize the same observation data to eliminate the dimensional differences between different features; S3, feature collection: select some features related to germination prediction in the observed data, and convert them into numerical features through label encoding method; S4. Dataset division: The cleaned and preprocessed dataset is divided into training set, validation set and test set. The training set is used to train the model, the validation set is used to adjust the model parameters and select hyperparameters for tuning, and the test set is used to evaluate the performance of the model. S5. Build a random forest model. Initialize the model using the machine learning library scikit-learn. Set the numerical features as hyperparameters. S6, model training training data and building decision trees, each decision tree is trained based on a random subset of the training data; S7. Model evaluation and tuning Use the validation set to evaluate the performance of the model and use grid search, random search, or Bayesian optimization to adjust the model's hyperparameters; S8. Prediction: Prepare new crop planting data, and use the trained random forest model to predict the new data through the same S2 data cleaning and S3 feature collection steps as the training data.

2. According to claim 1, a crop planting germination model prediction method based on the Internet of Things is characterized in that In the S1, some features related to germination prediction in the observed data are regularly replaced, and converted into numerical features through a label encoding method.

3. According to claim 1, a crop planting germination model prediction method based on the Internet of Things is characterized in that In the S1, observation data of germination organisms are collected periodically.

Images