Artificial intelligence red date water and fertilizer real-time monitoring method and system

Abstract

The invention discloses an artificial intelligence red date water and fertilizer real-time monitoring method and system, and the method comprises the steps: carrying out the precise recognition and segmentation of a red date tree body through a machine vision technology, extracting the growth state characteristic parameters of the red date tree body, such as the size of a tree crown, the color of leaves, and the number of fruits, and taking the parameters as important input data for predicting the requirements of red date water and fertilizer; a deep learning algorithm is adopted to construct a red date water and fertilizer demand intelligent prediction model, and multi-source heterogeneous data such as red date tree body growth state characteristics, real-time meteorological data, soil moisture content monitoring data and growth cycle stages are fused to obtain a water and fertilizer demand prediction model; and a data fusion algorithm is utilized to perform feature extraction and feature selection on red date tree body growth state features, real-time meteorological data, soil moisture content monitoring data and the like, redundant and noise data are removed, and the precision and generalization ability of a water and fertilizer demand prediction model are improved.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to an artificial intelligence-based real-time monitoring method and system for water and fertilizer of red dates. Background Art

[0002] In the real-time monitoring system of water and fertilizer for red dates based on artificial intelligence, how to accurately predict the real-time demand for water and fertilizer of red dates trees is a key technical issue. First, it is necessary to accurately identify and segment the red date trees through machine vision technology to extract the growth status characteristics of the trees, including key parameters such as crown size, leaf color, and number of fruits. Secondly, it is necessary to comprehensively analyze real-time meteorological data, such as light intensity, temperature and humidity, wind speed and direction, etc. These environmental factors have a direct impact on the transpiration and water and nutrient requirements of red date trees. Furthermore, soil moisture monitoring is another key point in predicting the water and fertilizer requirements of red dates. It is necessary to deploy soil moisture and nutrient sensors to collect soil water potential and nutrient content data at different depths in real time. Finally, it is also necessary to consider the growth cycle stage of the red date tree, such as flowering period, fruit setting period, and expansion period. The demand for water and fertilizer varies greatly at different stages. Based on the above multi-source heterogeneous data, building an intelligent prediction model for the water and fertilizer demand of red dates is a complex technical problem. It requires the use of artificial intelligence algorithms such as deep learning and data fusion to train a reliable and robust water and fertilizer demand prediction model to provide real-time basis for precise irrigation and fertilization of red dates. The generalization ability and adaptability of the model are also a technical challenge that needs to be overcome urgently. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide an artificial intelligence real-time monitoring method and system for water and fertilizer of red dates.

[0004] To achieve the above object, the present invention adopts the following technical solution:

[0005] An artificial intelligence-based real-time monitoring method for water and fertilizer of red dates, comprising:

[0006] Obtaining characteristic parameters of the growth status of jujube trees;

[0007] Obtain meteorological data and soil moisture monitoring data;

[0008] According to the characteristic parameters of the growth status of the jujube tree, meteorological data and soil moisture monitoring data, the growth stage of the jujube tree is determined;

[0009] The growth characteristics of jujube trees, real-time meteorological data, and soil moisture monitoring data were integrated;

[0010] Build a water and fertilizer demand prediction model based on the integrated characteristics;

[0011] Based on the water and fertilizer demand prediction model, irrigation and fertilization decisions are generated and automated control is achieved.

[0012] Preferably, the jujube trees are identified and segmented, and characteristic parameters of the jujube tree growth status such as crown size, leaf color, and fruit quantity are extracted.

[0013] Preferably, the meteorological data include light intensity, temperature and humidity, wind speed and direction; the soil moisture monitoring data include soil water potential and nutrient content data at different depths.

[0014] As a preferred method, the growth cycle stage of the jujube tree is analyzed, the current growth stage is determined based on the jujube growth cycle model, and the standard values ​​of the water and fertilizer requirements of the jujube tree at each growth stage are obtained; wherein the growth stages include: flowering period, fruit setting period and swelling period.

[0015] The present invention can also provide an artificial intelligence red date water and fertilizer real-time monitoring system, comprising:

[0016] The first processing module is used to obtain the characteristic parameters of the growth state of the jujube tree;

[0017] The second processing module is used to obtain meteorological data and soil moisture monitoring data;

[0018] The third processing module is used to determine the growth stage of the jujube tree according to the characteristic parameters of the jujube tree growth state, meteorological data and soil moisture monitoring data;

[0019] The fourth processing module is used to integrate the growth status characteristics of the jujube tree, real-time meteorological data, and soil moisture monitoring data;

[0020] The fifth processing module is used to build a water and fertilizer demand prediction model based on the fused characteristics;

[0021] The sixth processing module is used to generate irrigation and fertilization decisions and realize automatic control according to the water and fertilizer demand prediction model.

[0022] Preferably, the first processing module identifies and segments the jujube tree, and extracts characteristic parameters of the jujube tree growth status such as crown size, leaf color, and fruit quantity.

[0023] Preferably, the meteorological data include light intensity, temperature and humidity, wind speed and direction; the soil moisture monitoring data include soil water potential and nutrient content data at different depths.

[0024] Preferably, the third processing module analyzes the growth cycle stage of the jujube tree, determines the current growth stage based on the jujube growth cycle model, and obtains the standard values ​​of the jujube tree's demand for water and fertilizer at each growth stage; wherein the growth stages include: flowering period, fruit setting period and swelling period.

[0025] The present invention uses machine vision technology to accurately identify and segment the jujube tree, and extracts characteristic parameters of growth status such as crown size, leaf color, and fruit quantity. At the same time, real-time meteorological data and soil moisture monitoring data are obtained to analyze the impact of environmental factors on the water and fertilizer requirements of the jujube tree. Combined with the jujube growth cycle model, the standard requirements for water and fertilizer in the current growth stage are judged. A deep learning algorithm is used to fuse multi-source heterogeneous data to construct an intelligent prediction model for the water and fertilizer requirements of jujube. Features are extracted and selected through a data fusion algorithm to remove redundancy and noise, thereby improving the accuracy and generalization ability of the model. Finally, the trained model is applied to predict the water and fertilizer requirements of the jujube tree in real time, generate precise irrigation and fertilization decisions, and realize precise management of water and fertilizer of jujube through an automated control system. The present invention realizes the intelligent management of water and fertilizer of jujube, can effectively improve the yield and quality of jujube, and has important application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 The present invention is a flow chart of the method for real-time monitoring of water and fertilizer of red dates. DETAILED DESCRIPTION

[0027] In order to further understand the content of the present invention, the present invention is described in detail in conjunction with the accompanying drawings and embodiments. The present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It is understood that the specific embodiments described herein are only used to explain the relevant inventions, rather than to limit the invention. It is also necessary to explain that, for ease of description, only the parts related to the invention are shown in the accompanying drawings.

[0028] Embodiment 1:

[0029] like Figure 1 As shown, an artificial intelligence red date water and fertilizer real-time monitoring system according to an embodiment of the present invention may specifically include:

[0030] Step S101, accurately identifying and segmenting the jujube tree through machine vision technology, extracting the characteristic parameters of the jujube tree growth status such as crown size, leaf color, and fruit quantity as important input data for predicting the water and fertilizer requirements of the jujube.

[0031] The image data of a jujube tree is obtained, and the image data is segmented by an image segmentation algorithm to obtain an image of a single jujube tree; for the segmented single jujube tree image, the growth state characteristic parameters of the single jujube tree are extracted, and the growth state characteristic parameters include crown size, leaf color and fruit quantity, wherein the crown size is obtained by calculating the pixel area of ​​the crown region after segmentation, the leaf color is obtained by performing color statistical analysis on the leaf pixel points in the crown region, and the fruit quantity is obtained by identifying and counting the fruit targets in the crown region using a target detection algorithm; the extracted growth state characteristic parameters are input into a pre-established jujube tree water and fertilizer demand prediction model, and the water and fertilizer demand prediction model is used to calculate the fruit quantity. to the water and fertilizer demand of the jujube tree; if the water and fertilizer demand is higher than the preset fertilization threshold, an irrigation instruction is sent to the irrigation control system, and the irrigation duration and fertilization amount of the jujube tree are controlled according to the water and fertilizer demand; if the water and fertilizer demand does not exceed the preset fertilization threshold, there is no need to send an irrigation instruction to the irrigation control system; at different stages of the growth of the jujube tree, according to the growth state characteristic parameters of the jujube tree at different growth stages, the weights of each characteristic parameter in the water and fertilizer demand prediction model of the jujube tree are dynamically adjusted; the growth state changes of the jujube tree are continuously monitored, and when it is detected that the growth state characteristic parameters of the jujube tree have changed significantly, the water and fertilizer demand prediction result is updated, and the irrigation duration and fertilization amount of the irrigation control system are dynamically adjusted.

[0032] Specifically, obtaining image data of jujube trees can be achieved in a variety of ways. For example, using a drone equipped with a high-definition camera to take aerial photos of the jujube garden from the air to obtain image data covering the entire jujube garden. Cameras can also be deployed on the ground to take images of jujube trees at regular intervals. Handheld cameras can also be used to take photos of specific jujube trees at close range. Aerial photography can quickly obtain data over a large range, ground deployment can achieve long-term continuous monitoring, and handheld photography is suitable for fine data collection of specific trees. These methods can be flexibly selected and combined according to actual needs. Image segmentation algorithm is the process of segmenting an image into multiple regions with specific meanings. For jujube tree image segmentation, the purpose is to separate a single jujube tree from the background (such as soil, sky, etc.). Commonly used image segmentation algorithms include threshold-based segmentation, edge detection segmentation, region growing segmentation, and deep learning-based segmentation. For example, a color-based threshold segmentation method can be selected to identify the green part (mainly the crown and leaves) in the image as the jujube tree, and the other parts as the background. A semantic segmentation algorithm based on deep learning, such as U-Net, can also be used to train a model that can accurately identify the jujube tree body to achieve a more accurate segmentation effect. Accurate segmentation is the basis for subsequent feature extraction. After segmentation, the growth status feature parameters need to be extracted from the single jujube tree body image. The crown size can be obtained by calculating the pixel area of ​​the segmented crown area. For example, if the segmented crown area contains 10,000 pixels, the crown size is 10,000 pixels. The leaf color can be obtained by performing color statistical analysis on the leaf pixels in the crown area. For example, the number of green pixels, the number of red pixels, and the number of yellow pixels are counted, and then the ratio of these color pixels is calculated to determine whether the color of the leaf is dark green, light green, or yellowish. The number of fruits can be obtained by identifying and counting the fruit targets in the crown area using the target detection algorithm. For example, using the YOLOv5 target detection algorithm, if 50 jujube fruits are identified in the crown area, the number of fruits is 50. These feature parameters quantitatively describe the growth status of the jujube tree body. The water and fertilizer demand prediction model for jujube trees can predict the water and fertilizer demand of jujube trees based on the extracted growth status characteristic parameters. This model can be established based on expert experience or trained by machine learning algorithms based on a large amount of data. For example, a simple linear model can be established with crown size, leaf color and fruit quantity as input features, water and fertilizer demand as output, and model parameters fit by historical data. The model can predict how much water and fertilizer the current tree needs. Compare the predicted water and fertilizer demand with the preset fertilization threshold. Assume that the preset fertilization threshold is 10, the predicted water demand is 15, and the fertilizer demand is 12. Since both the water demand and fertilizer demand are higher than the fertilization threshold, an irrigation instruction is sent to the irrigation control system.The irrigation control system controls the irrigation duration and fertilizer application amount of the jujube tree according to the predicted water demand 15 and fertilizer demand 12. For example, the irrigation system is controlled to open for 15 minutes and 12 units of fertilizer are applied. If the predicted water demand is 8 and the fertilizer demand is 5, neither of which exceeds the fertilization threshold, there is no need to perform irrigation and fertilization operations. At different stages of the growth of the jujube tree, its growth state and water and fertilizer requirements will change. For example, in the seedling stage, the crown is small, the leaf color is lighter, the number of fruits is zero, and the main demand is water to promote growth; while in the fruiting stage, the crown is larger, the leaf color is darker, the number of fruits is larger, and the demand for water and fertilizer is relatively high. Therefore, it is necessary to dynamically adjust the weights of each characteristic parameter in the water and fertilizer demand prediction model of the jujube tree according to the growth state characteristic parameters of the jujube tree at different growth stages. For example, in the seedling stage, the weight of the leaf color can be increased and the weight of the number of fruits can be reduced; while in the fruiting stage, the weight of the number of fruits can be increased, and it can be judged whether specific fertilizers need to be supplemented based on the leaf color. Continuously monitor the changes in the growth status of the jujube tree. When the growth status characteristic parameters of the jujube tree are detected to have obvious changes, such as a sudden decrease in the size of the crown, yellowing of the leaves, and a decrease in the number of fruits, it indicates that the tree may have abnormal conditions, such as pests and diseases or lack of water and fertilizer. At this time, it is necessary to update the water and fertilizer demand forecast results and dynamically adjust the irrigation time and fertilizer application amount of the irrigation control system. For example, if the leaves are detected to be yellowing, it means that there may be a lack of nitrogen fertilizer, and the amount of nitrogen fertilizer application needs to be increased. Through dynamic adjustment, the growth changes of the tree can be responded to in a timely manner to ensure the healthy growth of the tree.

[0033] Step S102, obtain real-time meteorological data of the jujube planting area, including parameters such as light intensity, temperature and humidity, wind speed and direction, analyze the impact of these environmental factors on the transpiration and water and nutrient requirements of the jujube trees, and the correlation between changes in meteorological data and changes in the water and fertilizer requirements of the trees, and incorporate them into the water and fertilizer demand prediction model.

[0034] By deploying meteorological sensors in the jujube planting area, real-time meteorological data including light intensity, temperature and humidity, wind speed and direction are obtained. The acquired meteorological data is transmitted to the data processing center, and the data is preprocessed such as cleaning and normalization. According to the preprocessed meteorological data, the influence of environmental factors such as light intensity, temperature and humidity, wind speed and direction on the transpiration of jujube trees is analyzed, and the changing law of tree transpiration rate under different meteorological conditions is obtained. According to the change of tree transpiration rate, the water demand of jujube trees under different meteorological conditions is calculated, and its nutrient demand is estimated in combination with the growth and development stage of the tree. The time series analysis method is used to study the correlation between the change of meteorological data and the change of water and fertilizer demand of the tree within a certain time range, and a mathematical model between meteorological factors and water and fertilizer demand is established. The mathematical model between meteorological factors and water and fertilizer demand is incorporated into the machine learning algorithm for training to obtain a water and fertilizer demand prediction model. The model takes meteorological data as input and outputs the predicted value of water and fertilizer demand of jujube trees in the future. Based on the output results of the water and fertilizer demand prediction model, a precise irrigation and fertilization plan is formulated for the jujube planting area. Through the intelligent irrigation system and fertilization device, timely and quantitative water and fertilizer management is achieved to improve the yield and quality of jujube.

[0035] Specifically, meteorological data is crucial to the growth of jujube trees, and accurate acquisition of meteorological data can guide the refined management of jujube planting. Meteorological sensors are deployed in jujube planting areas to monitor a variety of meteorological parameters in real time. For example, multiple meteorological stations are installed in the jujube garden, each equipped with a light intensity sensor, a temperature and humidity sensor, a wind speed and direction sensor, etc., which can collect data such as light intensity, air temperature, air humidity, wind speed, and wind direction in different areas of the jujube garden in real time. The collected raw meteorological data usually needs to be preprocessed before it can be used for subsequent analysis. For example, there may be outliers in the light intensity data collected by a sensor, and these outliers need to be removed through data cleaning to ensure the accuracy of the data. In addition, different types of meteorological data, such as light intensity and temperature, may have very different numerical ranges, and they need to be normalized to convert the data to the same numerical range for subsequent model training and prediction. After the meteorological data preprocessing is completed, it is necessary to analyze the impact of meteorological factors on the transpiration of jujube trees. For example, when the light intensity is high, the temperature is high, the air humidity is low, and the wind speed is high, the transpiration of the jujube tree will be enhanced, and the water loss will be faster. On the contrary, when the light intensity is low, the temperature is low, the air humidity is high, and the wind speed is low, the transpiration of the jujube tree will be weakened, and the water loss will be slow. By analyzing the changing law of the transpiration rate of the tree under different meteorological conditions, the relationship between meteorological factors and transpiration rate can be established. Assuming that in clear weather, the light intensity is 10,000 lux, the temperature is 30 degrees Celsius, the humidity is 50%, and the wind speed is 2 meters per second, the transpiration rate of the jujube tree is measured to be 10 grams per hour. On cloudy days, when the light intensity is 2,000 lux, the temperature is 20 degrees Celsius, the humidity is 80%, and the wind speed is 0.5 meters per second, the transpiration rate of the jujube tree is measured to be 2 grams per hour. According to the changes in the transpiration rate of the tree, the water demand of the jujube tree can be calculated. For example, based on the transpiration rate measured above, it can be inferred that the jujube tree needs to add 10 grams of water per hour in sunny weather, while only 2 grams of water per hour is needed in cloudy weather. In addition, it is necessary to estimate the nutrient demand of the tree in combination with its growth and development stage. For example, during the flowering period and fruit expansion period of the jujube tree, the demand for nutrients is large, and more fertilizers need to be added. In order to establish a mathematical model between meteorological factors and water and fertilizer demand, a time series analysis method can be used. For example, collect meteorological data and corresponding tree water and fertilizer demand data over a period of time in the past, and analyze the correlation between them. Suppose the meteorological data of the past month shows that the higher the temperature, the greater the water demand of the jujube tree. Then a linear regression model can be established to use temperature to predict the water demand of the jujube tree. By incorporating the established mathematical model into the machine learning algorithm for training, a more accurate water and fertilizer demand prediction model can be obtained.For example, machine learning algorithms such as support vector machines and neural networks can be used to train historical data to obtain a model that can predict the water and fertilizer requirements of jujube trees in the future based on meteorological data. According to the output results of the water and fertilizer demand prediction model, an accurate irrigation and fertilization plan can be formulated. For example, if the model predicts that the water demand of jujube trees will be large in the next three days, irrigation can be carried out regularly and quantitatively through the intelligent irrigation system. At the same time, according to the nutrient demand predicted by the model, fertilizer can be applied accurately through the fertilization device, thereby improving the yield and quality of jujube.

[0036] Step S103, placing soil moisture and nutrient sensors in the jujube planting soil, collecting soil water potential and nutrient content data at different depths in real time, judging the soil moisture and nutrient consumption rate based on the soil moisture monitoring data, estimating the soil water and fertilizer supply capacity, and providing a basis for predicting the water and fertilizer demand of the tree.

[0037] According to the characteristics of the soil environment for jujube planting, appropriate soil moisture and nutrient sensors are selected and deployed at different depths to form a three-dimensional monitoring network. The sensors collect soil moisture and nutrient data at each depth in real time and upload them to the data processing center through the wireless communication module. In the data processing center, the collected soil moisture and nutrient data are pre-processed by cleaning and normalization to obtain standardized time series data. Using time series analysis algorithms, such as ARIMA model or LSTM neural network, a prediction model for soil moisture and nutrient consumption rate is established, and the model parameters are trained according to historical data. Through machine learning regression algorithms, such as support vector machine regression or random forest regression, a relationship model between soil moisture, nutrients and soil supply capacity is established. The predicted results of soil moisture and nutrient consumption rate are input into the relationship model of soil supply capacity to estimate the current soil water and fertilizer supply capacity. Taking into account the tree growth stage and environmental factors, the water and fertilizer demand prediction model is used, combined with the estimated results of soil water and fertilizer supply capacity, to predict the water and fertilizer demand of jujube trees, providing a decision-making basis for precision irrigation and fertilization.

[0038] Specifically, soil moisture and nutrients are key factors in the growth and development of red dates. In order to achieve precise irrigation and fertilization, it is necessary to monitor the soil water and fertilizer conditions in real time and precisely regulate them according to the needs of the red date trees. First, it is necessary to select appropriate sensors according to the soil characteristics of the red date planting area. For example, in sandy soil, you can choose a dielectric constant soil moisture sensor with fast response speed and high accuracy, and an ion selective electrode nutrient sensor with a wide measurement range. In clay, you can choose a soil moisture sensor with a larger measurement range and an optical nutrient sensor with stronger anti-interference ability. Assuming that in a red date plantation, the soil texture is loam, and FDR soil moisture sensor and EC-5 soil nutrient sensor are selected. The depth of sensor layout needs to consider the distribution of red date roots. The roots of red date trees are mainly distributed in the soil layer of 0-60 cm, so the sensors can be laid out at three depths of 10 cm, 30 cm and 50 cm to form a three-dimensional monitoring network. Set 3 repetitions at each depth to facilitate data correction and improve monitoring accuracy. The data collected by the sensor is uploaded to the data processing center in real time through a wireless communication module, such as LoRa or NB-IoT. Assume that the soil moisture data collected by a sensor at a depth of 10 cm is 15%, 20% at a depth of 30 cm, and 25% at a depth of 50 cm, and the nutrient content is 100 ppm, 80 ppm, and 60 ppm respectively. In the data processing center, the collected data must first be cleaned to remove outliers and erroneous data. For example, if a sensor fails and the data is abnormally high or low, it needs to be removed. Then, the data is normalized to convert data of different dimensions into a unified dimension for subsequent analysis. For example, the soil moisture data is converted to a value between 0-1, and the nutrient data is converted to a value between 0-100. Assume that after cleaning and normalization, the soil moisture data at a depth of 10 cm is 0.6 and the nutrient data is 60. Using the time series analysis algorithm, the consumption rate of soil moisture and nutrients can be predicted. For example, the ARIMA model is used to fit the soil moisture data in the past period of time to predict the trend of soil moisture in the future period of time. Suppose that according to the ARIMA model, the soil moisture at a depth of 10 cm will drop by 0.1, that is, to 0.5, in the next 24 hours. The same method can be used to predict the rate of nutrient consumption. Suppose the nutrient content is predicted to drop by 10ppm, that is, to 50ppm. Through machine learning regression algorithms, a relationship model between soil moisture, nutrients and soil supply capacity can be established. For example, using the support vector machine regression algorithm, the model is trained based on historical data to learn the nonlinear relationship between soil moisture, nutrients and soil supply capacity. Suppose the model predicts that when the soil moisture content is 0.5 and the nutrient content is 50ppm, the soil supply capacity is medium.Comprehensively consider the tree growth stage and environmental factors. For example, the current jujube tree is in the fruit expansion period and requires more water and nutrients. At the same time, considering the current high temperature and vigorous transpiration, the tree's demand for water is further increased. According to the water and fertilizer demand prediction model, combined with the estimated results of the soil water and fertilizer supply capacity, the water and fertilizer demand of the jujube tree can be predicted. For example, the model predicts that the jujube tree will need to supplement 10 liters of water and 500 grams of nutrients in the next 24 hours. Based on this, a precise irrigation and fertilization plan is formulated, and the timed and quantitative water and fertilizer management is achieved through the intelligent irrigation system and fertilization device, ultimately improving the yield and quality of jujube.

[0039] Step S104, comprehensively analyze the growth cycle stage of the jujube tree, judge the current growth stage, such as flowering stage, fruit setting stage, swelling stage, etc. according to the jujube growth cycle model, obtain the standard values ​​of water and fertilizer requirements of the jujube tree at each growth stage, as a baseline reference for predicting the water and fertilizer requirements of the tree.

[0040] According to the growth cycle model of the jujube tree, the start and end time nodes and duration data of each growth stage are obtained to establish a time law library of the growth stage. Computer vision technology is used to collect images of organs such as branches, leaves, flowers, and fruits of the jujube tree, and the morphological characteristic parameters of each organ are extracted through image segmentation algorithms. The extracted morphological characteristic parameters of the tree organs are matched with the pre-established morphological characteristic models of the organs in each growth stage to determine the current growth stage of the jujube tree. According to the current growth stage determined, the standard value of the tree water and fertilizer demand in this stage is queried from the water and fertilizer demand standard value database as a baseline reference for subsequent prediction of the tree water and fertilizer demand. If the current time is in the transition period between two growth stages, the weighted average of the water and fertilizer demand standard values ​​of the two stages is used to obtain the baseline reference value of the water and fertilizer demand in the transition period. The soil moisture, temperature and humidity data of the jujube tree planting area are collected through environmental sensors, and the tree water and fertilizer demand prediction model is trained by combining the water and fertilizer demand standard values ​​of the current growth stage with machine learning algorithms such as support vector machines. Using the trained water and fertilizer demand prediction model, the current water and fertilizer demand of the jujube tree can be predicted in real time according to the tree's current growth stage and environmental factor data, guiding precise irrigation and fertilization.

[0041] Specifically, the growth and development of jujube trees show obvious periodic changes, and understanding their growth cycle is crucial for precise water and fertilizer management. According to the phenological period of jujube, the growth cycle of jujube trees can be divided into several key stages, including budding, leafing, flowering, fruiting, fruit expansion, maturity and leaf fall. And record the start and end time nodes and duration of each stage. For example, the budding period generally starts around mid-March in spring and lasts for about 15 days; the leafing period is after the budding period and lasts for about 20 days. By analogy, a growth stage time law library is constructed. In order to determine the current growth stage of the jujube tree, computer vision technology can be used. For example, at different growth stages of the jujube tree, images of its branches, leaves, flowers and fruits are collected respectively. For example, during the flowering period, images of jujube flowers are taken, and during the fruit expansion period, images of jujube fruits are taken. Using image segmentation algorithms, the target organs (such as flowers and fruits) in the image are segmented from the background. After that, the morphological characteristic parameters of the segmented organs are extracted, such as the length, width, and area of ​​the leaves, the diameter and color of the flowers, the diameter and color of the fruits, etc. These extracted characteristic parameters are then matched with the pre-established morphological characteristic models of the organs at each growth stage. For example, the characteristic model of the flowering period may include information such as the diameter and color of the flowers. By calculating the similarity between the extracted characteristic parameters and the models of each stage, the growth stage of the current jujube tree can be determined. Assuming that the currently extracted fruit morphological characteristic parameters have the highest similarity with the fruit expansion period model, it can be determined that the current jujube tree is in the fruit expansion period. If the current time is in the transition period between two growth stages, such as between the flowering period and the fruit setting period, the water and fertilizer demand standard values ​​of the two stages can be weighted averaged to obtain the baseline reference value of the water and fertilizer demand in the transition period. For example, assuming that the standard value of water and fertilizer demand during the flowering period is 10, and the standard value of water and fertilizer demand during the fruiting period is 15, the current time is in the transition period between the flowering period and the fruiting period, and is closer to the fruiting period, then the weight of the fruiting period can be set to be larger, for example, 0.7, and the weight of the flowering period is 0.3. The final calculated baseline reference value of water and fertilizer demand during the transition period is 10*0.3+15*0.7=13.5. In order to predict the water and fertilizer demand of the jujube tree, it is necessary to collect environmental data of the planting area, such as soil moisture, temperature, humidity, etc. These data can be collected in real time by sensors deployed in the field. Combined with the current growth stage of the jujube tree and the standard value of water and fertilizer demand at this stage, for example, more water and nutrients are required during the fruit expansion period, these data can be used as input features, and machine learning algorithms, such as support vector machines, can be used to train the tree water and fertilizer demand prediction model. Through model training, the complex relationship between environmental factors, growth stages and water and fertilizer demand can be learned. When making predictions, the current growth stage, soil moisture, temperature, humidity and other data are input into the trained model to obtain the predicted value of the current water and fertilizer demand of the jujube tree.For example, if the jujube trees are currently in the fruit expansion period, the soil moisture is suitable, the temperature is high, and the humidity is low, the model may predict a higher water and fertilizer demand, such as 20 liters of water and 0.5 kg of fertilizer per tree per day. Accurate prediction results can guide fruit farmers to carry out precise irrigation and fertilization to avoid excessive or insufficient irrigation, thereby improving the utilization efficiency of water and fertilizer, and ultimately improving the yield and quality of jujubes.

[0042] The soil moisture, temperature and humidity data of the jujube tree planting area are collected through environmental sensors, and the tree water and fertilizer demand prediction model is trained using a machine learning algorithm based on the standard values ​​of water and fertilizer requirements in the current growth stage.

[0043] Environmental sensors are used to collect environmental data such as soil moisture, temperature and humidity in the jujube tree planting area, and the collected data is transmitted to the data processing center. According to the growth stage of the jujube tree, the water and fertilizer demand standard values ​​of the corresponding stage are obtained from the preset water and fertilizer demand standard value database. The environmental data and water and fertilizer demand standard values ​​collected by the environmental sensor are input into the machine learning model, and the model is trained to obtain the trained water and fertilizer demand prediction model. The trained water and fertilizer demand prediction model is used in combination with the real-time collected environmental data to predict the water and fertilizer demand of the jujube tree in the current growth stage. According to the predicted water and fertilizer demand, the irrigation system and fertilization system are controlled to achieve precise water and fertilizer management in the jujube tree planting area. In the process of water and fertilizer management, environmental data and jujube tree growth data are continuously collected and fed back to the machine learning model to continuously optimize and update the model. Through the continuously optimized water and fertilizer demand prediction model, the irrigation and fertilization strategies are dynamically adjusted to achieve precise water and fertilizer management of the entire planting cycle of the jujube tree, and improve the yield and quality of the jujube.

[0044] Specifically, the precise water and fertilizer management of jujube trees requires the integration of environmental perception, growth stage identification, machine learning and other technologies, ultimately achieving efficient utilization of water and fertilizer resources and improving the yield and quality of jujubes. First, it is necessary to understand the environmental conditions of the jujube planting area in real time. Sensor network technology can be used to deploy soil moisture sensors, temperature and humidity sensors, light sensors, etc. in the field to form an environmental monitoring network. The soil moisture sensor can monitor the moisture content of the soil. For example, the current soil moisture content is 20%, which is lower than the suitable range for jujube growth. The temperature and humidity sensor can monitor the air temperature and humidity. For example, the current temperature is 25 degrees Celsius and the humidity is 60%. The light sensor can monitor the light intensity. For example, the current light intensity is 10,000 lux. The data collected by these sensors are transmitted to the data processing center via a wireless network to provide basic data for subsequent water and fertilizer demand forecasts. Secondly, it is necessary to determine the current growth stage of the jujube tree. The growth cycle of the jujube tree can be divided into budding period, flowering period, fruiting period, swelling period, maturity period, etc. The duration and organ morphological characteristics of each stage are different. For example, the flowering period lasts about 15 days, with a large number of flowers and bright colors. Computer vision technology can be used to collect images of jujube trees, and the morphological characteristic parameters of tree organs, such as the number, size, and color of flowers, can be extracted through image recognition algorithms. Then, the extracted parameters are matched with the pre-established organ morphological characteristic models of each growth stage to determine the current growth stage of the jujube tree. For example, if a large number of blooming jujube flowers are detected in the image, it is judged that the tree is currently in the flowering period. Thirdly, it is necessary to obtain the standard values ​​of water and fertilizer requirements of jujube trees at different growth stages. These standard values ​​can refer to the experience of agricultural experts or experimental data and be stored in the database. For example, during the flowering period, the jujube tree has a high demand for water, requiring about 10 liters of water per plant per day, and a relatively low demand for fertilizers, for example, a fertilizer with a nitrogen, phosphorus, and potassium ratio of 1:0.5:1 is required, and the amount of fertilizer applied each time is 100 grams. Next, a water and fertilizer demand prediction model is to be constructed. The environmental data and the standard values ​​of water and fertilizer requirements for the corresponding growth stage are input into the machine learning model for training. Common machine learning algorithms include support vector machines and neural networks. For example, the soil moisture, temperature and humidity, light intensity, and standard values ​​of water and fertilizer requirements during the flowering period are input into the support vector machine model for training to obtain a model that can predict the water and fertilizer requirements during the flowering period based on environmental data. Finally, the trained prediction model is used for real-time prediction. For example, the current jujube tree is in the flowering period, the soil moisture is 20%, the temperature is 25 degrees Celsius, the humidity is 60%, and the light intensity is 10,000 lux. By inputting these data into the trained support vector machine model, it can be predicted that the current jujube tree needs 8 liters of water and 80 grams of fertilizer. According to the prediction results, the irrigation system and fertilization system can be controlled for precise water and fertilizer management.For example, the irrigation system is controlled to provide 8 liters of water to each jujube tree, and the fertilization system is controlled to apply 80 grams of fertilizer with a nitrogen, phosphorus and potassium ratio of 1:0.5:1. In practical applications, it is necessary to continuously collect environmental data and jujube tree growth data, and feed them back into the machine learning model to continuously optimize the model and improve the prediction accuracy. In this way, the irrigation and fertilization strategies can be dynamically adjusted according to the growth status of the jujube tree and environmental changes, and accurate water and fertilizer management of the entire planting cycle of the jujube tree can be achieved, ultimately improving the yield and quality of jujubes. For example, if it is found that there is a large deviation between the predicted water and fertilizer demand and the actual demand for several consecutive days, it is necessary to readjust the model parameters or select a more appropriate machine learning algorithm to improve the prediction accuracy of the model.

[0045] Step S105, using a deep learning algorithm to build an intelligent prediction model for the water and fertilizer requirements of red dates, integrating multi-source heterogeneous data such as the growth status characteristics of red date trees, real-time meteorological data, soil moisture monitoring data, and growth cycle stages, to train a reliable and robust water and fertilizer demand prediction model.

[0046] The growth status characteristic data of jujube trees are obtained, including tree age, tree height, crown diameter, leaf area index, etc. Meteorological data such as temperature, humidity, light intensity, wind speed, etc. are collected in real time through IoT sensors, and soil moisture monitoring data, including soil moisture, temperature, conductivity, etc. are obtained. According to the growth cycle stage of jujube, the multi-source heterogeneous data obtained are preprocessed and feature extracted. Data fusion technology is used to unify and align data from different sources and formats to construct a training data set for water and fertilizer demand prediction of jujube. Deep learning algorithms such as long short-term memory network (LSTM) or convolutional neural network (CNN) are used to construct an intelligent prediction model for water and fertilizer demand of jujube. The fused multi-source heterogeneous data are input into the model for training. The model performance is optimized by adjusting the model parameters and hyperparameters. During the model training process, the model is evaluated by cross-validation and other methods. The prediction accuracy and generalization ability of the model are evaluated by calculating indicators such as mean square error (MSE) and mean absolute error (MAE). It is continuously iterated and optimized until a reliable and robust prediction model is obtained. The trained intelligent prediction model for water and fertilizer demand of red dates is deployed in the production environment, and the growth status characteristics, meteorological data and soil moisture monitoring data of red dates are obtained in real time. The model is input for prediction to obtain the water and fertilizer demand of red dates in the future. According to the prediction results, combined with the growth cycle stage of red dates and the knowledge of agronomic experts, an accurate water and fertilizer management plan is formulated. Through the intelligent irrigation and fertilization system, the accurate management of water and fertilizer of red dates is achieved, the efficiency of fertilizer and water utilization is improved, and the healthy growth, high yield and high quality of red dates are promoted. A feedback optimization mechanism for the intelligent prediction of water and fertilizer demand of red dates is established, and the actual water and fertilizer usage is compared with the predicted value. The prediction model is continuously optimized and updated to improve the adaptability and robustness of the model, providing continuous and reliable decision-making support for the accurate management of water and fertilizer of red dates.

[0047] Specifically, the growth status characteristic data of the jujube tree, such as tree age, tree height, crown diameter, leaf area index, etc., can reflect the growth status and size of the tree, and are an important basis for predicting the demand for water and fertilizer. For example, a 10-year-old jujube tree with a height of 3 meters and a crown diameter of 4 meters, and a 5-year-old jujube tree with a height of 2 meters and a crown diameter of 2 meters, even at the same growth stage, their demand for water and fertilizer is definitely different. The older the tree, the larger the tree, and the greater the demand for water and fertilizer. The leaf area index can reflect the number and size of the leaves of the tree. The leaves are the main organs for photosynthesis. The larger the leaf area, the stronger the photosynthesis, and the higher the demand for water and fertilizer. These data can be collected manually or automatically using sensors. IoT sensors can collect meteorological data in real time, such as temperature, humidity, light intensity, wind speed, etc., which will affect the transpiration and photosynthesis of the jujube tree, and then affect the demand for water and fertilizer. For example, in a hot and dry environment, the transpiration of the jujube tree is enhanced, and the demand for water will increase. In an environment with insufficient light, photosynthesis weakens and the demand for fertilizer will decrease. Soil moisture monitoring data, including soil moisture, temperature, conductivity, etc., can directly reflect the soil's ability to supply water and fertilizer. Low soil moisture content indicates that water needs to be replenished in time; too low or too high soil temperature will affect the absorption capacity of the root system, thereby affecting the demand for water and fertilizer. Soil conductivity can reflect the salt content in the soil. Excessive salt will inhibit root growth and affect water and fertilizer absorption. The growth cycle of red dates can be roughly divided into several stages, including germination, flowering, fruit setting, swelling, maturity and dormancy. The demand for water and fertilizer for red date trees varies greatly at different growth stages. For example, during the flowering and fruit setting periods, red date trees need a lot of nutrients to support the development of flowers and fruits, and the demand for water and fertilizer is relatively high. In the dormant period, the tree basically stops growing and the demand for water and fertilizer is very low. For different growth cycle stages, it is necessary to preprocess and extract features from the collected multi-source heterogeneous data. For example, smooth the time series data to remove noise and outliers; extract features from image data, such as color, texture, shape, etc. Data fusion technology can unify and align data from different sources and formats to build a training data set for predicting the water and fertilizer requirements of red dates. For example, tree growth status feature data, meteorological data, soil moisture monitoring data, etc. can be aligned according to the time dimension to form a data set containing multiple features for model training. Deep learning algorithms, such as long short-term memory networks (LSTM) or convolutional neural networks (CNN), can be used to build an intelligent prediction model for the water and fertilizer requirements of red dates. LSTM networks are suitable for processing time series data and can capture the temporal dependencies between data. For example, the meteorological data and soil moisture data of the past few days will affect the water and fertilizer requirements of red dates today.CNN network is suitable for processing image data and can extract features from images. For example, by analyzing the image of the leaves of the jujube tree, the growth status of the tree can be judged, and then the water and fertilizer requirements can be predicted. During the model training process, it is necessary to use cross-validation and other methods to evaluate the model, calculate indicators such as mean square error (MSE) and mean absolute error (MAE), and evaluate the prediction accuracy and generalization ability of the model. For example, the data set is divided into a training set and a test set, the model is trained with the training set, and the prediction performance of the model is evaluated with the test set. By continuously adjusting the model parameters and hyperparameters, the model performance is optimized until a reliable and robust prediction model is obtained. The trained model is deployed to the production environment, data is obtained in real time, and the model is input for prediction to obtain the water and fertilizer requirements of the jujube in the future. For example, the model can be deployed on a cloud server, and data can be collected in real time through the Internet of Things device, and the data is transmitted to the server. The model on the server makes predictions and feeds the prediction results back to the irrigation and fertilization system. According to the prediction results, combined with the growth cycle stage of the jujube and the knowledge of agronomic experts, an accurate water and fertilizer management plan can be formulated. For example, if it is predicted that there will be more rainfall in the next few days, the amount of irrigation can be reduced; if it is predicted that the jujube tree is about to enter the swelling period, the amount of fertilizer applied can be increased. A feedback optimization mechanism is established to compare the actual water and fertilizer usage with the predicted values, continuously optimize and update the prediction model, and improve the adaptability and robustness of the model. For example, the actual amount of irrigation and fertilizer application can be recorded and compared with the model's predicted values. If a large deviation is found, the cause needs to be analyzed and the model adjusted to improve the prediction accuracy.

[0048] Step S106, using data fusion algorithm, extracting and selecting features from the growth status characteristics of jujube trees, real-time meteorological data, soil moisture monitoring data, etc., removing redundant and noisy data, and improving the accuracy and generalization ability of the water and fertilizer demand prediction model.

[0049] The method comprises the following steps: obtaining characteristic data of the growth state of jujube trees, real-time meteorological data and soil moisture monitoring data, preprocessing the acquired data, removing abnormal values ​​and missing values ​​therein, and obtaining a preprocessed data set; extracting features of various types of data according to the preprocessed data set by using the principal component analysis method, extracting key features that can reflect the growth state of jujube trees, meteorological conditions and soil moisture conditions, and obtaining a characteristic data set; removing redundant features and noise features therein through correlation analysis and redundancy calculation, and selecting a preferred feature subset that is closely related to the prediction of water and fertilizer requirements; inputting the preferred feature subset into a support vector machine model for training, and optimizing the hyperparameters of the support vector machine model through grid search and cross-validation methods. parameters to obtain a water and fertilizer demand prediction model with excellent performance; using the growth status characteristics of jujube trees, real-time meteorological data and soil moisture monitoring data as input, passing them into the water and fertilizer demand prediction model to obtain the water and fertilizer demand prediction results of the jujube trees; according to the water and fertilizer demand prediction results, combined with the growth status characteristics of jujube trees, real-time meteorological data and soil moisture monitoring data, dynamically adjust the irrigation and fertilization plans to achieve precise and intelligent water and fertilizer management of jujube trees; continuously collect the growth status characteristics of jujube trees, real-time meteorological data and soil moisture monitoring data, regularly retrain and optimize the water and fertilizer demand prediction model, continuously improve the accuracy and generalization ability of the water and fertilizer demand prediction model, and provide long-term intelligent decision-making support for water and fertilizer management of jujube trees.

[0050] Specifically, the growth status characteristic data of the jujube tree is obtained. For example, the height of the tree is measured by a sensor. The height of one jujube tree is 3 meters and the height of the other is 2.5 meters. The crown diameter can be estimated by drone aerial images. For example, the crown diameter of one tree is 4 meters and the crown diameter of the other is 3 meters. The leaf area index can be measured using a handheld leaf area meter. For example, the leaf area indexes of the two trees are measured to be 3.5 and 2.8 respectively. Real-time meteorological data can be obtained through a meteorological station. For example, the temperature at a certain moment is 25 degrees Celsius, the humidity is 60%, and the light intensity is 1000 watts / square meter. Soil moisture monitoring data can be obtained through sensors buried in the soil. For example, the soil moisture content is 20% and the soil temperature is 22 degrees Celsius. The acquired data is preprocessed to remove outliers and missing values. For example, if the temperature measurement value at a certain moment is -50 degrees Celsius, this is obviously an outlier and needs to be removed. If the soil moisture content data at a certain moment is missing, the data at the previous moment or the next moment can be used for interpolation and supplementation. The preprocessed data set will be used for subsequent feature extraction and model training. The principal component analysis method is used to extract features from various types of data. For example, principal component analysis is performed on growth status feature data such as tree height, crown diameter, and leaf area index to extract principal components that can reflect tree size and growth status. Principal component analysis is performed on meteorological data such as temperature, humidity, and light intensity to extract principal components that can reflect climatic conditions. Principal component analysis is performed on soil moisture monitoring data such as soil moisture, temperature, and electrical conductivity to extract principal components that can reflect soil fertility. These principal components will constitute the feature data set. Through correlation analysis and redundancy calculation, redundant features and noise features are removed, and the preferred feature subset that is closely related to water and fertilizer demand prediction is selected. For example, if a high correlation is found between tree height and crown diameter, only one of the features can be retained. If it is found that a feature has almost no relationship with water and fertilizer demand, it can be regarded as a noise feature and removed. The final preferred feature subset will be used for model training. The preferred feature subset is input into the support vector machine model for training. For example, the radial basis kernel function can be used as the kernel function of the support vector machine. The hyperparameters of the support vector machine are optimized by grid search and cross-validation methods, for example, the penalty coefficient C and the kernel function parameter γ are optimized. Finally, a water and fertilizer demand prediction model with excellent performance is obtained. The growth status characteristics of the jujube tree, real-time meteorological data and soil moisture monitoring data are used as inputs, which are passed into the water and fertilizer demand prediction model to obtain the water and fertilizer demand prediction results of the jujube tree. For example, the input data is 3 meters in height, 4 meters in crown diameter, 3.5 in leaf area index, 25 degrees Celsius in temperature, 60% in humidity, and 20% in soil moisture content. The prediction model outputs the water and fertilizer demand for the next week, for example, 10 liters of water and 0.5 kg of fertilizer are needed.According to the results of water and fertilizer demand forecast, combined with the growth status characteristics of jujube trees, real-time meteorological data and soil moisture monitoring data, the irrigation and fertilization schemes are adjusted dynamically. For example, if the rainfall is predicted to be large in the next week, the irrigation amount can be reduced. If the soil fertility is high, the amount of fertilizer can be reduced. By dynamically adjusting the irrigation and fertilization schemes, the water and fertilizer management of jujube trees can be precise and intelligent. The growth status characteristics of jujube trees, real-time meteorological data and soil moisture monitoring data are continuously collected, and the water and fertilizer demand prediction model is retrained and optimized regularly. For example, data is collected every other month, and the model is retrained with new data to improve the accuracy and generalization ability of the model. Doing so can enable the model to better adapt to environmental changes and tree growth changes, and provide long-term intelligent decision-making support for water and fertilizer management of jujube trees.

[0051] Step S107, using the trained intelligent prediction model for water and fertilizer demand of red dates, predicting the demand for water and fertilizer of red dates in real time, generating precise irrigation and fertilization decisions for red dates, and realizing precise management of water and fertilizer for red dates through an automated control system, thereby improving the yield and quality of red dates and realizing intelligent management of water and fertilizer for red dates.

[0052] The growth status data of jujube trees, including soil moisture, nutrient content, meteorological conditions, etc., are obtained as inputs to the water and fertilizer demand prediction model; based on the acquired growth status data of jujube trees, the trained jujube water and fertilizer demand intelligent prediction model is used to predict the demand for water and nutrients of the jujube trees in the current growth stage in real time; based on the output results of the jujube water and fertilizer demand prediction model, combined with the set jujube yield and quality targets, a precise irrigation and fertilization decision plan for jujubes is generated, including irrigation time, irrigation amount, fertilization time, fertilization amount, etc.; the generated precise irrigation and fertilization decision plan for jujubes is transmitted to the automatic control system to control the irrigation equipment and fertilization The equipment performs precise irrigation and fertilization operations according to the decision-making plan; continuously obtains the growth status data of jujube trees during the growth process of jujube, predicts the changes in the water and fertilizer demand of jujube in real time, dynamically adjusts the irrigation and fertilization decisions, and realizes the dynamic optimization of the water and fertilizer management of jujube; uses machine learning algorithms to analyze historical jujube water and fertilizer management data, explores the correlation between water and fertilizer management and jujube yield and quality, and continuously optimizes the jujube water and fertilizer demand prediction model and decision-making algorithm; builds an intelligent platform for jujube water and fertilizer management, integrating jujube water and fertilizer demand prediction, precise irrigation and fertilization decision-making, and automatic control, to realize the intelligence and precision of the entire process of jujube water and fertilizer management.

[0053] Specifically, the growth status data of jujube trees is the basis for predicting water and fertilizer requirements. These data include soil moisture, nutrient content, and meteorological conditions. For example, soil moisture sensors can be used to monitor soil moisture in real time, soil nutrient rapid testers can be used to determine the content of nutrients such as nitrogen, phosphorus, and potassium in the soil, and meteorological data such as temperature, humidity, and light can be obtained from meteorological stations. Together, these data constitute a comprehensive description of the growth environment of jujube trees, providing data support for subsequent precise water and fertilizer management. By obtaining these data, we can better understand the environmental conditions required for the growth of jujube trees, so as to carry out targeted water and fertilizer management. The trained jujube water and fertilizer demand intelligent prediction model is used to predict the water and fertilizer demand of jujube trees in real time. The model is trained based on a large amount of historical data and machine learning algorithms, and can accurately predict its demand for water and nutrients based on the current jujube tree growth status data. For example, the model can predict how much water the jujube tree needs to supplement based on the current soil moisture being below a certain threshold; at the same time, according to the growth stage of the jujube tree and the soil nutrient content, it can predict how much nitrogen, phosphorus, potassium and other nutrients need to be supplemented. The advantage of doing so is that it can avoid excessive or insufficient irrigation and fertilization, thereby improving water and fertilizer utilization efficiency and saving resources. Combined with the red date yield and quality targets, formulate an accurate irrigation and fertilization decision plan. The output results of the water and fertilizer demand prediction model are only instructive and need to be adjusted according to actual conditions. For example, if the red date yield target is expected to be high this year, it is necessary to increase the supply of water and fertilizer appropriately based on the model prediction; if higher red date quality is pursued, it is necessary to adjust the proportion of nutrients such as nitrogen, phosphorus, and potassium to achieve the best nutritional balance. When formulating a decision plan, it is also necessary to consider specific operational details such as irrigation time, irrigation amount, fertilization time, and fertilization amount to ensure the executability of the plan. Only in this way can the prediction results be converted into actual action plans to guide precise irrigation and fertilization. The generated precise irrigation and fertilization decision plan needs to be executed through an automated control system. The automated control system can receive instructions in the decision plan and control the irrigation equipment and fertilization equipment to operate according to the predetermined plan. For example, the system can control the switch and flow of the irrigation valve to achieve timed and quantitative irrigation; it can also control the operation of the fertilizer applicator to accurately deliver fertilizer to the soil around the red date tree. The application of automated control systems can reduce manual operations, improve the efficiency and accuracy of water and fertilizer management, and reduce labor costs. The growth state of jujube trees changes dynamically, so it is necessary to continuously monitor and adjust the irrigation and fertilization plan according to actual conditions. For example, if there is continuous rainfall and the soil moisture is high, it is necessary to reduce the amount of irrigation or even suspend irrigation; if the jujube tree is found to have nutrient deficiency symptoms, it is necessary to supplement the corresponding nutrients in time. Continuous monitoring and dynamic adjustment can ensure that the jujube tree is always in the best growth state and improve yield and quality. By analyzing historical jujube water and fertilizer management data through machine learning algorithms, the water and fertilizer demand prediction model and decision-making algorithm can be continuously optimized.For example, data on the yield, quality, water and fertilizer input of red dates in the past few years can be collected, and the correlation between these data can be analyzed using machine learning algorithms to find out the key factors affecting the yield and quality of red dates, and incorporated into the prediction model and decision-making algorithm. Continuous optimization can continuously improve the prediction accuracy of the model and the scientific nature of decision-making, and ultimately achieve intelligent and precise water and fertilizer management of red dates. An intelligent platform for water and fertilizer management of red dates is built to integrate functions such as water and fertilizer demand forecasting, precise irrigation and fertilization decision-making, and automatic control into a unified platform to achieve intelligent and precise water and fertilizer management of red dates. For example, the platform can display information such as soil moisture, nutrient content, and meteorological data in real time, and automatically generate irrigation and fertilization plans based on these data, and transmit the plans to the automatic control system for execution. Such a platform can facilitate managers to monitor the growth status and water and fertilizer management of red date trees in real time, discover and solve problems in a timely manner, and improve management efficiency.

[0054] Embodiment 2:

[0055] The present invention can also provide an artificial intelligence red date water and fertilizer real-time monitoring system, comprising:

[0056] The first processing module is used to obtain the characteristic parameters of the growth state of the jujube tree;

[0057] The second processing module is used to obtain meteorological data and soil moisture monitoring data;

[0058] The third processing module is used to determine the growth stage of the jujube tree according to the characteristic parameters of the jujube tree growth state, meteorological data and soil moisture monitoring data;

[0059] The fourth processing module is used to integrate the growth status characteristics of the jujube tree, real-time meteorological data, and soil moisture monitoring data;

[0060] The fifth processing module is used to build a water and fertilizer demand prediction model based on the fused characteristics;

[0061] The sixth processing module is used to generate irrigation and fertilization decisions and realize automatic control according to the water and fertilizer demand prediction model.

[0062] As an implementation method of the embodiment of the present invention, the first processing module identifies and segments the jujube tree, and extracts the characteristic parameters of the jujube tree growth state such as the crown size, leaf color, and fruit quantity.

[0063] As an implementation method of an embodiment of the present invention, the meteorological data includes light intensity, temperature and humidity, wind speed and direction; the soil moisture monitoring data includes soil water potential and nutrient content data at different depths.

[0064] As an implementation mode of an embodiment of the present invention, the third processing module analyzes the growth cycle stage of the jujube tree, determines the current growth stage according to the jujube growth cycle model, and obtains the standard values ​​of the water and fertilizer requirements of the jujube tree at each growth stage; wherein the growth stages include: flowering period, fruit setting period and swelling period.

[0065] The above is only a specific implementation method of this specification. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the system, module and unit described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here. It should be understood that the protection scope of this specification is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this specification, and these modifications or replacements should be covered within the protection scope of this specification.

Claims

1. An artificial intelligence-based real-time monitoring method for water and fertilizer of red dates, characterized in that: include: Obtaining characteristic parameters of the growth status of jujube trees; Obtain meteorological data and soil moisture monitoring data; According to the characteristic parameters of the growth status of the jujube tree, meteorological data and soil moisture monitoring data, the growth stage of the jujube tree is determined; The growth characteristics of jujube trees, real-time meteorological data, and soil moisture monitoring data were integrated; Build a water and fertilizer demand prediction model based on the integrated characteristics; Based on the water and fertilizer demand prediction model, irrigation and fertilization decisions are generated and automated control is achieved.

2. The artificial intelligence jujube water and fertilizer real-time monitoring method according to claim 1, characterized in that: The jujube trees are identified and segmented, and the characteristic parameters of the jujube tree growth status such as crown size, leaf color and fruit number are extracted.

3. The artificial intelligence jujube water and fertilizer real-time monitoring method as claimed in claim 2, characterized in that: Meteorological data include light intensity, temperature and humidity, wind speed and direction; soil moisture monitoring data include soil water potential and nutrient content data at different depths.

4. The artificial intelligence jujube water and fertilizer real-time monitoring method as claimed in claim 3 is characterized in that: Analyze the growth cycle stage of the jujube tree, determine the current growth stage according to the jujube growth cycle model, and obtain the standard values ​​of water and fertilizer requirements of the jujube tree at each growth stage; among them, the growth stages include: flowering period, fruit setting period and swelling period.

5. An artificial intelligence red date water and fertilizer real-time monitoring system, characterized in that: include: The first processing module is used to obtain the characteristic parameters of the growth state of the jujube tree; The second processing module is used to obtain meteorological data and soil moisture monitoring data; The third processing module is used to determine the growth stage of the jujube tree according to the characteristic parameters of the jujube tree growth state, meteorological data and soil moisture monitoring data; The fourth processing module is used to integrate the growth status characteristics of the jujube tree, real-time meteorological data, and soil moisture monitoring data; The fifth processing module is used to build a water and fertilizer demand prediction model based on the fused characteristics; The sixth processing module is used to generate irrigation and fertilization decisions and realize automatic control according to the water and fertilizer demand prediction model.

6. The artificial intelligence red date water and fertilizer real-time monitoring system as claimed in claim 5, characterized in that: The first processing module identifies and segments the jujube tree, and extracts the characteristic parameters of the jujube tree growth status such as crown size, leaf color, and fruit quantity.

7. The artificial intelligence jujube water and fertilizer real-time monitoring system according to claim 6, characterized in that: Meteorological data include light intensity, temperature and humidity, wind speed and direction; soil moisture monitoring data include soil water potential and nutrient content data at different depths.

8. The artificial intelligence jujube water and fertilizer real-time monitoring system as claimed in claim 7 is characterized in that: The third processing module analyzes the growth cycle stage of the jujube tree, determines the current growth stage according to the jujube growth cycle model, and obtains the standard values ​​of water and fertilizer requirements of the jujube tree at each growth stage; wherein the growth stage includes: flowering stage, fruit setting stage and swelling stage.

Images