Citrus seedling water stress monitoring system, method, computer device and medium

Abstract

The application provides a citrus seedling water stress monitoring system, method, computer equipment and medium, the system comprises a host computer, a first wireless data transmission terminal, a second wireless data transmission terminal and a soil temperature and humidity sensor, the first wireless data transmission terminal is a USB to LoRa wireless data transmission terminal, the second wireless data transmission terminal is a RS485 to LoRa wireless data transmission terminal, the host computer is connected with the first wireless data transmission terminal in a serial communication mode, the first wireless data transmission terminal establishes communication with the second wireless data transmission terminal in a broadcast communication mode, and the second wireless data transmission terminal is connected with the soil temperature and humidity sensor.The application can realize unmanned and wireless monitoring of seedling changes and states on the basis of nondestructive monitoring, and through data viewing, the abnormal data area can be accurately positioned, the occurrence of soil diseases can be prevented, irrigation can be more accurately carried out, seedling can be more scientifically cultivated, and root water gradient experiments on seedlings are facilitated.

Description

Technical Field

[0001] This invention relates to a citrus seedling water stress monitoring system, method, computer equipment, and medium, belonging to the fields of seedling cultivation technology and algorithm application. Background Technology

[0002] The mainstream citrus seedling cultivation is gradually shifting from growing seedlings in the ground to cultivating them using seedling bags and nutrient soil. Currently, most seedling cultivation environments in China involve filling seedling bags with soil and burying the bags in the ground. Once the seedlings reach a certain stage of growth, they are pulled out of the bags for transplanting or sale. There are no greenhouses, and the overall environment is no different from direct soil cultivation. The advantage is that the seedling bags protect the roots during transplanting and save labor. For example, contact with the soil allows for longer water retention. The disadvantage is that the seedlings are in a wild environment and may encounter natural disasters such as heavy rainfall, as well as pests and diseases hidden in the soil. Overseas seedling cultivation is more modern, mainly reflected in seedling greenhouses, including both closed and open types. Seedling bags are placed on the ground or on specific shelves. The bags are filled with nutrient soil, which is more breathable and nutrient-rich. The disadvantage is poor water retention, requiring irrigation. However, this also results in more abundant water and oxygen for the roots. Overall, seedlings grown in greenhouses are less susceptible to disease and have stronger, more robust roots.

[0003] Research literature on waterlogging and drought damage to citrus seedlings is relatively limited, and currently, there are no relevant electronic systems or modern equipment to intervene in this crucial aspect of improving seedling quality. The main problems and shortcomings include the lack of clear boundaries and specific data regarding drought and waterlogging damage; and the difficulty in timely and early observation of seedling changes in unsuitable environments. Citrus seedlings inherently possess a certain degree of resistance to the environment, and short-term waterlogging and drought damage may not cause obvious changes in their phenotypes. Seedlings that appear normal on the surface may already be in a state of resistance. Only by using a chlorophyll meter and digging into the soil to observe the roots can one see changes in their sensitivity to the environment. These changes are actions taken to resist the environment, rather than a loss of resistance. For the large number of seedlings in a nursery, these two inspection methods are clearly impractical, and digging into the soil can cause damage to the roots. In addition, irrigation decisions are still based on the traditional method of observing whether the leaves are curled or shrunken to determine whether there is a lack of water. If there is too much water, treatment can only be carried out when obvious problems appear on the surface of the seedlings. By the time either of these situations occurs, the seedlings have already lost their ability to resist the environment and have entered a damaged state. Both of these situations illustrate that traditional irrigation methods cannot provide a comfortable environment for citrus seedlings in the long run. Environmental intervention can only be carried out after the seedlings give a signal, at which point the seedlings can no longer maintain a normal and healthy state.

[0004] Traditional seedling cultivation methods may suffer from several problems, including: mild drought leading to slow plant growth, affecting subsequent tree development and fruit quality, and weakened roots; mild waterlogging causing excessive root growth, resulting in thin, fragile root systems; insufficient oxygen leading to root rot and foul odor over time, eventually killing the plant; and high temperature and humidity in the soil increasing susceptibility to diseases and pests such as leaf miners, damping-off, and blight. Therefore, monitoring and pest control during the seedling cultivation process is crucial for future large-scale seedling bases, higher seedling quality, and the further realization of "digital agriculture" and "precision agriculture." Summary of the Invention

[0005] The first objective of this invention is to overcome the shortcomings and deficiencies of the prior art and propose a citrus seedling water stress monitoring system. This system can monitor the environment of the seedling roots, acquire and record data. For large-scale seedling bases, it can achieve unmanned and wireless monitoring of seedling changes and status based on non-destructive monitoring. At the same time, by viewing the data, it can accurately locate areas with abnormal data, which helps to prevent soil diseases, irrigate with higher precision, raise seedlings more scientifically, and conduct root water gradient experiments on seedlings.

[0006] The second objective of this invention is to provide a method for monitoring water stress in citrus seedlings.

[0007] A third objective of this invention is to provide a computer device.

[0008] A fourth objective of this invention is to provide a computer-readable storage medium.

[0009] The first objective of this invention can be achieved by adopting the following technical solution:

[0010] A citrus seedling water stress monitoring system includes a host computer, a first wireless data transmission terminal, a second wireless data transmission terminal, and a soil temperature and humidity sensor. The first wireless data transmission terminal is a USB-to-LoRa wireless data transmission terminal, and the second wireless data transmission terminal is an RS485-to-LoRa wireless data transmission terminal. The host computer is connected to the first wireless data transmission terminal via serial communication, and the first wireless data transmission terminal establishes communication with the second wireless data transmission terminal via broadcast communication. The second wireless data transmission terminal is connected to the soil temperature and humidity sensor.

[0011] The soil temperature and humidity sensor is used to collect soil environmental information of the seedling bags where the citrus seedlings are located.

[0012] The host computer is used to periodically acquire and record data from soil temperature and humidity sensors; receive selection commands input by the user; if the selection command is an evaluation command, it acquires data from the past first preset time period, processes the data to generate a first monitoring dataset, and puts the first monitoring dataset into a trained multilayer perceptron model to obtain an evaluation result of seedling growth over the past first preset time period; if the selection command is a pest prevention command, it acquires data from the past second preset time period, processes the data to generate a second monitoring dataset, and puts the second monitoring dataset into a trained multilayer perceptron model to obtain a predicted growth state of seedlings after maintaining the environment of the past second preset time period for a first preset time period; wherein, the first preset time period is longer than the second preset time period.

[0013] Furthermore, the multilayer perceptron model includes an input layer, two hidden layers, and an output layer, with the first preset time being 7 days and the second preset time being 1 day;

[0014] The input layer has 504 neurons, corresponding to 504 data points for every 7 days; the hidden layers have 168 neurons and 7 neurons, corresponding to hours and days respectively; the output layer has 3 neurons, corresponding to 3 categories; the activation function in the hidden layers is the sigmoid function, the activation function in the output layers is the softmax function, and the BatchNorm1d function is used between the layers.

[0015] Furthermore, the soil temperature and humidity sensor is vertically inserted into the soil environment of the seedling bag where the citrus seedling is located, so that the probe is completely submerged in the soil, and the insertion position is 4cm away from the main stem of the citrus seedling.

[0016] Furthermore, it also includes an extension antenna, with the SMA interface of the first and second wireless data transmission terminals connected to the extension antenna to achieve data transmission and reception within a range of 100m.

[0017] The second objective of this invention can be achieved by adopting the following technical solution:

[0018] A method for monitoring water stress in citrus seedlings, the method comprising:

[0019] Data from soil temperature and humidity sensors is acquired and recorded periodically.

[0020] Receive selection instructions from the user;

[0021] If the instruction is selected as the evaluation instruction, the data from the past first preset time period will be obtained, the data will be processed to generate the first monitoring dataset, and the first monitoring dataset will be put into the trained multilayer perceptron model to obtain the evaluation results of seedling growth in the past first preset time period.

[0022] If the selected instruction is the pest prevention instruction, the data from the past second preset time period is obtained, the data is processed, a second monitoring dataset is generated, and the second monitoring dataset is put into the trained multilayer perceptron model to obtain the predicted growth status of the seedlings after the environment of the past second preset time period is maintained for a first preset time period.

[0023] The first preset time is greater than the second preset time.

[0024] Furthermore, the method also includes:

[0025] A dataset of temperature and humidity data and seedling status is created. The dataset represents the various water content environments in which the roots are located within a first preset time period. The dataset includes a training set and a test set. The data and classification of the training set are derived from the original experimental data and simulations and combinations of the experimental data. The data and classification of the test set are derived from individual data from each day.

[0026] The data in the dataset is processed, and the processed dataset is fed into the multilayer perceptron model for training. After adjusting and optimizing the functions of each layer, the trained multilayer perceptron model is obtained.

[0027] Furthermore, the processing of the data in the dataset specifically includes:

[0028] Sort the values ​​in each data entry from largest to smallest;

[0029] For similar data, data from a single day of similar experiments are randomly extracted and synthesized to obtain new data for that type of experiment, and dissimilar data are mixed in proportion.

[0030] The original data is shifted up or down by a few small units to fill in the boundaries between different classes.

[0031] Furthermore, the acquisition of data from the soil temperature and humidity sensor specifically includes:

[0032] Set the query frames for each soil temperature and humidity sensor in sequence, initialize the dataframes for storing data from each soil temperature and humidity sensor, check if the serial port is open, send the query frames for each soil temperature and humidity sensor one by one, receive the returned information, transcode each one, calculate the temperature and humidity, obtain the current time, and write the time and temperature and humidity into the dataframes corresponding to each soil temperature and humidity sensor.

[0033] Furthermore, after feeding the first monitoring dataset into the trained multilayer perceptron model to obtain the evaluation results of seedling growth over a first preset time period, the process further includes:

[0034] If the assessment result indicates an unhealthy state, and the assessment result is drought, it is recommended that the first pre-set irrigation in the future increase the soil moisture content within a healthy range to allow the plants to replenish and store moisture. If the assessment result is waterlogging, it is recommended that the first pre-set irrigation in the future decrease the soil moisture content within a healthy range to slow down the tendency of the plant roots to continue to grow elongate, increase the radial growth tendency of the plant roots, and make the plant roots grow stronger.

[0035] The third objective of this invention can be achieved by adopting the following technical solution:

[0036] A computer device includes a processor and a memory for storing processor-executable programs, wherein when the processor executes the programs stored in the memory, it implements the above-described method for monitoring water stress in citrus seedlings.

[0037] The fourth objective of this invention can be achieved by adopting the following technical solution:

[0038] A computer-readable storage medium storing a program that, when executed by a processor, implements the above-described method for monitoring water stress in citrus seedlings.

[0039] The present invention has the following advantages over the prior art:

[0040] This invention, based on non-destructive monitoring, digitizes seedling monitoring during the seedling stage. It monitors the environment surrounding the seedling roots, acquires and records data, and enables unmanned, wireless monitoring of seedling changes and status in large-scale seedling bases. For new varieties, experiments on leaves and roots can be conducted, and the collected data can be used to determine the most suitable growth environment for that variety. Simultaneously, data analysis can aid in preventing soil-borne diseases, achieving more precise irrigation, more scientific seedling cultivation, and conducting root moisture gradient experiments on seedlings. By collecting and recording data, selecting assessment or prevention modules, processing the data, and inputting it into a self-built and trained MLP model, the corresponding seedling growth status (healthy, drought-damaged, and waterlogged) can be determined without any physical contact with the seedlings. Based on the growth status and current known information about the seedlings, environmental adjustments can be made in advance to prevent drought and waterlogging. After drought or waterlogging occurs, the seedling environment can be adjusted more rationally based on data and system results. With timely adjustments and prevention, seedlings can always be in an ideal growth environment during the seedling cultivation process, ensuring that the roots of the seedlings grow both thick and long, the main stem grows stronger, the possibility of disease is reduced, and preparation is made for subsequent transplanting, flowering, and fruiting. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0042] Figure 1 This is a block diagram of the citrus seedling water stress monitoring system according to Embodiment 1 of the present invention.

[0043] Figure 2 This is a schematic diagram of the citrus seedling water stress monitoring system of Embodiment 1 of the present invention.

[0044] Figure 3 This is a schematic diagram of the structure of the multilayer perceptron model in Embodiment 1 of the present invention.

[0045] Figure 4 This is a schematic diagram of the evaluation module and the damage prevention module of Embodiment 1 of the present invention.

[0046] Figure 5 This is a flowchart of the method for monitoring water stress in citrus seedlings according to Embodiment 2 of the present invention.

[0047] Figure 6 This is a structural block diagram of the computer device according to Embodiment 3 of the present invention. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0049] Example 1:

[0050] like Figure 1 and Figure 2 As shown in the figure, this embodiment provides a citrus seedling water stress monitoring system. The system includes a host computer, a first wireless data transmission terminal, a second wireless data transmission terminal, an extension antenna, a soil temperature and humidity sensor, a 220V power supply, and a DC12V1A power adapter.

[0051] The system includes a soil temperature and humidity sensor for collecting soil environmental information from the seedling bags containing citrus seedlings; a host computer for periodically acquiring and recording data from the soil temperature and humidity sensor; and a user-input selection command. If the selection command is an evaluation command, the system acquires data from the past first preset time period, processes the data to generate a first monitoring dataset, and inputs the first monitoring dataset into a trained multilayer perceptron model to obtain an evaluation result of seedling growth over the past first preset time period. If the selection command is a pest control command, the system acquires data from the past second preset time period, processes the data to generate a second monitoring dataset, and inputs the second monitoring dataset into a trained multilayer perceptron model to obtain a predicted growth status of the seedlings after maintaining the environment of the past second preset time period for a first preset time period. The first preset time period is 7 days, and the second preset time period is 1 day.

[0052] In this embodiment, the host computer is a PC, which connects to the first wireless data transmission terminal via serial communication through a CH430 driver. The host computer's USB interface provides 5V power to LoRa and serves as a communication interface. The host computer includes software such as Excel and PyCharm; the overall operating system includes environments such as Python and PyTorch; and a self-built and trained Multi-Layer Perceptron (MLP) model. The host computer has two main tasks: first, to communicate with lower-level devices such as soil temperature and humidity sensors to acquire, record, and save data to realize the monitoring function; and second, to act as the main implementer of the monitoring system, saving sensor data and running algorithms.

[0053] Furthermore, Excel and PyCharm, through the writing and running of Python programs, complete the data acquisition task. The specific process is as follows: query frames for each sensor are set sequentially; data frames for storing data from different sensors are initialized; the serial port is checked for openness; query frames from each sensor are sent sequentially at 0.2-second intervals; the returned information is received, transcoded sequentially, and the temperature and humidity are calculated; the current time is obtained; and the time, temperature, and humidity are written into the corresponding data frames for each sensor. After acquiring a certain amount of data or after a certain period of time, each data frame is named using the format of date + sensor serial number and written to an Excel spreadsheet. The monitoring system generates a new spreadsheet every 24 hours from startup.

[0054] like Figure 3As shown, the multilayer perceptron model in this embodiment includes an input layer, two hidden layers, and an output layer. The input layer has 504 neurons, corresponding to 504 data points for every 7 days; the hidden layers have 168 neurons and 7 neurons, corresponding to the number of hours and days, respectively; the output layer has 3 neurons, corresponding to 3 categories; the activation function in the hidden layers is the sigmoid function, the activation function in the output layers is the softmax function, and the BatchNorm1d function is used between the layers.

[0055] Furthermore, for seedlings, with fewer branches and leaves and in a growth stage, the roots play an irreplaceable role in terms of both nutrition and water. Ultimately, any impact on the roots during seedling cultivation is caused by changes in water content. Through experiments, a dataset of temperature, humidity, and seedling status was created. The resulting dataset has water content denoted as hi (i = 1, 2, ... 504), representing the root's environment at that water content over 20 minutes. The entire dataset represents the various water content environments the roots experienced over a 7-day period. The dataset includes a training set and a test set. The training set data and classifications are derived from original experimental data and simulations of experimental data, ensuring the classifications fully reflect real-world conditions. It contains 700 tensors. The test set data and... The classification is based on individual data from each day, with the classification results corresponding to the experimental group of each day. There are a total of 73 tensors, each with a size of 1*504, a stride of 35, and a learning rate of 0.0001. Ideally, after the network has been trained, it should be able to correctly classify the 73 data points in the test set that are for the purpose of preventing harm. The processed dataset is then fed into a multilayer perceptron model for training. After adjusting and optimizing the functions of each layer, a relatively ideal multilayer perceptron model is finally obtained. The model achieves an accuracy of over 99.5% on the training set and over 97.5% on the test set.

[0056] Furthermore, the experiment included three types of experiments: drought, flood, and health. The experiment lasted for 7 days. The resulting dataset underwent data processing, including mixing within the same class and mixing dissimilar classes in proportion. The final dataset contained a total of 773 data points (700 for the training set and 73 for the test set).

[0057] Furthermore, the data processing includes the following three parts: First, the values ​​in each data point are arranged from largest to smallest. Since the main trend of humidity change is from large to small, external factors such as irrigation during seedling cultivation can cause the values ​​to rise, making the overall arrangement of the curves more distinct. Second, data filling involves mixing data within the same category, randomly sampling and synthesizing data from a single day of the same experiment to obtain new data for that category. Dissimilar categories are mixed proportionally; healthy mixed data has 5 or 6 healthy days, while unhealthy mixed data has 1-4 healthy days. Third, the original data is shifted up or down by a few small units to fill in the boundaries between different categories. Through data processing, almost all non-extreme situations that might occur in reality are covered, resulting in a reasonable and sufficient dataset.

[0058] Furthermore, the assessment and prevention functions in the host computer can be divided into an assessment module and a prevention module. After the monitoring system is started, it automatically acquires and records data according to the time intervals and order set by the program, and automatically generates data files, waiting to be viewed or the assessment module or prevention module is run.

[0059] Furthermore, the host computer also includes debugging software for soil temperature and humidity sensors, wireless data transmission terminals, etc., before the monitoring system runs. After debugging, the monitoring system program is run, and the host computer automatically acquires and records data from each sensor at regular intervals. After a period of time, the generated file data is reviewed, and a visualization program is run to view the root environment of seedlings in various areas during this period. Combined with environmental changes during this period, such as temperature, precipitation, and disasters, irrigation strategies are adjusted to provide seedlings with a comfortable growth environment, allowing the overall growth of seedlings to be more ideal. In situations where natural disasters or other factors cause significant environmental changes and data fluctuations, real-time monitoring data can help adjust the environment as quickly as possible, restoring the soil to an ideal state and reducing pests and diseases caused by these changes. For example, high temperature and humidity can easily lead to leaf miners, damping-off, blight, and other pests and diseases. For newly planted varieties, experiments on the leaves and roots, combined with monitoring data, can determine the most suitable growth environment for the variety. Digital monitoring of the entire seedling process provides useful and real-time information, as well as data analysis and processing for problems that arise, ultimately creating a more ideal seedling environment.

[0060] like Figure 4As shown, the evaluation module requires a monitoring dataset of more than 7 days. After running, it acquires and processes the data from the past 7 days to generate the dataset required by the algorithm (the first monitoring dataset). This dataset is then fed into a self-built and trained multilayer perceptron model to obtain the evaluation results of seedling growth over the past 7 days. The evaluation results include drought damage, healthy seedlings, and waterlogging damage. The damage prevention module requires a monitoring dataset of more than 1 day. After running, it acquires and processes the data from the past day to generate the dataset required by the algorithm (the second monitoring dataset). This dataset is then fed into a self-built and trained multilayer perceptron model to obtain the seedling growth status after maintaining the environment of the past day for 7 days. The growth status includes drought damage, healthy seedlings, and waterlogging damage.

[0061] Further, select the assessment module, and the program will generate an assessment result. Examine the areas corresponding to the unhealthy state and review the data from the past 7 days. If the data deviates significantly from the healthy range, observe the seedling surface characteristics for any changes and treat them promptly. If the data does not clearly indicate a problem, the plant surface may still appear normal, but the internal structure has gradually lost its resistance to the environment. Adjust the environment as soon as possible before the seedling is damaged to restore a suitable environment for its growth. Simultaneously, in an unhealthy assessment state, future irrigation strategies can be slightly adjusted to promote plant recovery. If the assessment indicates drought, irrigation for the next week can slightly increase soil moisture within a healthy range to allow the plant to replenish its reserves. If the assessment indicates waterlogging, irrigation for the next week can slightly decrease soil moisture within a healthy range to slow the elongation of the plant's roots, increase radial root growth, resulting in stronger roots and easier oxygen access.

[0062] Furthermore, selecting the pest control module will generate a prediction result after the program runs. Examine the areas corresponding to unhealthy states, reviewing the data from the past day for any sudden numerical changes or other anomalies. Check the corresponding areas to see if any accidental environmental damage has occurred, and address any abnormal numerical conditions as quickly as possible. If handled promptly, the seedlings, unaffected by the environmental changes, will return to a suitable environment and continue growing in their ideal condition. For large seedling nurseries, accidents are often difficult to detect promptly; the monitoring system and pest control module can quickly alert managers.

[0063] In this embodiment, the first wireless data transmission terminal is a USB to LoRa wireless data transmission terminal, and the second wireless data transmission terminal is an RS485 to LoRa wireless data transmission terminal. Before installation, the baud rate, device ID, wireless channel, wireless transmission power, and rate level of the first and second wireless data transmission terminals need to be set to establish a stable wireless communication network and avoid device conflicts.

[0064] Furthermore, the SMA interfaces of the first and second wireless data transmission terminals are connected to the extension antenna, which can easily achieve stable data transmission and reception within a range of 100m.

[0065] Furthermore, the first wireless data transmission terminal establishes communication with the second wireless data transmission terminal via broadcast communication. The second wireless data transmission terminal is connected to the soil temperature and humidity sensor. The wireless data transmission terminals conduct network communication in the form of broadcast, that is, any terminal device sends data packets to other devices within the broadcast domain, and all devices choose whether to receive and whether to provide feedback according to different settings. It has the advantages of simple network equipment, simple maintenance, and low network deployment cost. The server does not need to send data to each client individually, and the server traffic load is extremely low. The disadvantage is that it cannot provide personalized services. For this system, this disadvantage can be ignored, and its accurate information exchange can be handled by the soil temperature and humidity sensor.

[0066] Furthermore, the DC port of the second wireless data transmission terminal is connected to a DC12V1A power adapter, which is connected to a 220V power supply; the RS485a and RS485b ports are connected to the communication lines corresponding to the soil temperature and humidity sensor, and the VCC and GND terminals are connected to the power supply terminals of the soil temperature and humidity sensor to power it; the second wireless data transmission terminal, the 220V power supply, and the DC12V1A power adapter should be placed away from humid environments to ensure safety.

[0067] In this embodiment, the soil temperature and humidity sensor is an RS485 soil temperature and humidity sensor, which is vertically inserted into the soil environment of the seedling bag where the citrus seedling is located, so that the probe is completely submerged in the soil. At the same time, the insertion position is 4cm away from the main stem of the citrus seedling. The data acquired by the soil temperature and humidity sensor includes temperature and humidity, which represent the soil environment information of the current probe location and the surrounding area. If it is in the seedling bag environment, under almost the same external environment, it can represent the environment around the seedling bag where the probe is located as the center. That is, one soil temperature and humidity sensor can monitor at least 9 seedlings. If it is planted on soil, it can monitor the space within a 1m radius.

[0068] Furthermore, multiple soil temperature and humidity sensors can be combined on a single RS485 bus to enable a LoRa wireless data transmission terminal to connect to multiple sensors. Based on the hardware limitations of the LoRa wireless data transmission terminal, it is recommended that a maximum of three sensors be connected to one LoRa wireless data transmission terminal. In addition, the RS485 communication standard allows for the setting of address codes for different devices. This can be configured on the host computer before operation, giving each different soil temperature and humidity sensor a unique identifier. This allows query frames broadcast from the host computer to be sent to each soil temperature and humidity sensor, and the soil temperature and humidity sensor can determine whether to respond based on the address code, thereby enabling individual communication between the host computer and each soil temperature and humidity sensor.

[0069] Example 2:

[0070] like Figure 5 As shown in the figure, this embodiment provides a method for monitoring water stress in citrus seedlings, which includes the following steps:

[0071] S501: Periodically acquire and record data from the soil temperature and humidity sensor.

[0072] S502. Receive the user's input selection instruction. If the selection instruction is an evaluation instruction, proceed to step S503. If the selection instruction is a prevention instruction, proceed to step S504.

[0073] S503. Obtain data from the past first preset time period, process the data to generate a first monitoring dataset, and put the first monitoring dataset into the trained multilayer perceptron model to obtain the evaluation results of seedling growth within the past first preset time period.

[0074] S503. Obtain data within the past second preset time period, process the data to generate a second monitoring dataset, and put the second monitoring dataset into the trained multilayer perceptron model to obtain the predicted growth status of the seedlings after maintaining the environment of the past second preset time period for a first preset time period.

[0075] The specific implementation process of each step of the citrus seedling water stress monitoring method in this embodiment can be found in the citrus seedling water stress monitoring system in Embodiment 1 above, and will not be repeated here.

[0076] Example 3:

[0077] This embodiment provides a computer device, such as... Figure 6As shown, it includes a processor 602, a memory, an input device 603, a display 504, and a network interface 605 connected via a device bus 601. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium 606 and internal memory 607. The non-volatile storage medium 606 stores operating devices, computer programs, and a database. The internal memory 607 provides an environment for the operation of the operating devices and computer programs in the non-volatile storage medium. When the processor 602 executes the computer program stored in the memory, it implements the citrus seedling water stress monitoring method of Embodiment 2 described above, as follows:

[0078] Data from soil temperature and humidity sensors is acquired and recorded periodically.

[0079] Receive selection instructions from the user;

[0080] If the instruction is selected as the evaluation instruction, the data from the past first preset time period will be obtained, the data will be processed to generate the first monitoring dataset, and the first monitoring dataset will be put into the trained multilayer perceptron model to obtain the evaluation results of seedling growth in the past first preset time period.

[0081] If the selected instruction is the pest prevention instruction, the data from the past second preset time period is obtained, the data is processed, a second monitoring dataset is generated, and the second monitoring dataset is put into the trained multilayer perceptron model to obtain the predicted growth status of the seedlings after the environment of the past second preset time period is maintained for a first preset time period.

[0082] The first preset time is greater than the second preset time.

[0083] Example 4:

[0084] This embodiment provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the citrus seedling water stress monitoring method of Embodiment 2 above, as follows:

[0085] Data from soil temperature and humidity sensors is acquired and recorded periodically.

[0086] Receive selection instructions from the user;

[0087] If the instruction is selected as the evaluation instruction, the data from the past first preset time period will be obtained, the data will be processed to generate the first monitoring dataset, and the first monitoring dataset will be put into the trained multilayer perceptron model to obtain the evaluation results of seedling growth in the past first preset time period.

[0088] If the selected instruction is the pest prevention instruction, the data from the past second preset time period is obtained, the data is processed, a second monitoring dataset is generated, and the second monitoring dataset is put into the trained multilayer perceptron model to obtain the predicted growth status of the seedlings after the environment of the past second preset time period is maintained for a first preset time period.

[0089] The first preset time is greater than the second preset time.

[0090] It should be noted that the computer-readable storage medium in this embodiment can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor device, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0091] In this embodiment, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used or combined with an instruction execution device, apparatus, or device. In this embodiment, the computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying a computer-readable program. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The computer-readable signal medium can also be any computer-readable storage medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use or combined with an instruction execution device, apparatus, or device. The computer program contained on the computer-readable storage medium can be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination thereof.

[0092] The computer-readable storage medium described above can be used to write computer programs for executing this embodiment in one or more programming languages ​​or combinations thereof. These programming languages ​​include object-oriented programming languages—such as Java, Python, and C++—and conventional procedural programming languages—such as C or similar programming languages. The program can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0093] In summary, this invention, based on non-destructive monitoring, achieves digital monitoring of seedlings during the seedling cultivation process. It monitors the environment surrounding the seedling roots, acquires and records data, and enables unmanned, wireless monitoring of seedling changes and status in large-scale seedling bases. For new varieties, experiments on leaves and roots can be conducted, and the collected data can be used to determine the most suitable growth environment for that variety. Simultaneously, data analysis can aid in preventing soil-borne diseases, achieving more precise irrigation, more scientific seedling cultivation, and conducting root moisture gradient experiments on seedlings. By collecting and recording data, selecting assessment or prevention modules, processing the data, and inputting it into a self-built and trained MLP model, the corresponding seedling growth status (healthy, drought-damaged, and waterlogged) can be determined without any physical contact with the seedlings. Based on the growth status and current known information about the seedlings, environmental adjustments can be made in advance to prevent drought and waterlogging. After drought or waterlogging occurs, the seedling environment can be adjusted more rationally based on data and system results. With timely adjustments and prevention, seedlings can always be in an ideal growth environment during the seedling cultivation process, ensuring that the roots of the seedlings grow both thick and long, the main stem grows stronger, the possibility of disease is reduced, and preparation is made for subsequent transplanting, flowering, and fruiting.

[0094] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A citrus seedling water stress monitoring system, characterized in that, The system includes a host computer, a first wireless data transmission terminal, a second wireless data transmission terminal, and a soil temperature and humidity sensor. The first wireless data transmission terminal is a USB-to-LoRa wireless data transmission terminal, and the second wireless data transmission terminal is an RS485-to-LoRa wireless data transmission terminal. The host computer is connected to the first wireless data transmission terminal via serial communication, and the first wireless data transmission terminal establishes communication with the second wireless data transmission terminal via broadcast communication. The second wireless data transmission terminal is connected to the soil temperature and humidity sensor. The soil temperature and humidity sensor is used to collect soil environmental information of the seedling bags where the citrus seedlings are located. The soil temperature and humidity sensor is vertically inserted into the soil environment of the seedling bags where the citrus seedlings are located, so that the probe is completely submerged in the soil, and the insertion position is 4cm away from the main stem of the citrus seedling. The host computer is used to periodically acquire and record data from soil temperature and humidity sensors; receive selection commands input by the user; if the selection command is an evaluation command, it acquires data from the past first preset time period, processes the data to generate a first monitoring dataset, and inputs the first monitoring dataset into a trained multilayer perceptron model to obtain an evaluation result of seedling growth over the past first preset time period; if the selection command is a pest control command, it acquires data from the past second preset time period, processes the data to generate a second monitoring dataset, and inputs the second monitoring dataset into a trained multilayer perceptron model to obtain an evaluation result of seedling growth over the past second preset time period. The predicted growth status of seedlings after maintaining the environment for a first preset time is determined. The multilayer perceptron model includes an input layer, two hidden layers, and an output layer. The first preset time is 7 days, and the second preset time is 1 day. The input layer has 504 neurons, corresponding to 504 data points every 7 days. The hidden layers have 168 neurons and 7 neurons, corresponding to hours and days respectively. The output layer has 3 neurons, corresponding to 3 categories. The activation function in the hidden layers is the sigmoid function, and the activation function in the output layers is the softmax function. A BatchNorm1d function is used between the layers.

2. The citrus seedling water stress monitoring system according to claim 1, characterized in that, It also includes an extension antenna, and the SMA interface of the first and second wireless data transmission terminals is connected to the extension antenna to realize data transmission and reception within a range of 100m.

3. A method for monitoring water stress in citrus seedlings, implemented based on the citrus seedling water stress monitoring system according to any one of claims 1-2, characterized in that, The method includes: Data from soil temperature and humidity sensors is acquired and recorded periodically. Receive selection instructions from the user; If the instruction is selected as the evaluation instruction, the data from the past first preset time period will be obtained, the data will be processed to generate the first monitoring dataset, and the first monitoring dataset will be put into the trained multilayer perceptron model to obtain the evaluation results of seedling growth in the past first preset time period. If the selected instruction is the pest prevention instruction, the data from the past second preset time period is obtained, the data is processed, a second monitoring dataset is generated, and the second monitoring dataset is put into the trained multilayer perceptron model to obtain the predicted growth status of the seedlings after the environment of the past second preset time period is maintained for a first preset time period.

4. The method for monitoring water stress in citrus seedlings according to claim 3, characterized in that, The method further includes: A dataset of temperature and humidity data and seedling status is created. The dataset represents the various water content environments in which the roots are located within a first preset time period. The dataset includes a training set and a test set. The data and classification of the training set are derived from the original experimental data and simulations and combinations of the experimental data. The data and classification of the test set are derived from individual data from each day. The data in the dataset is processed, and the processed dataset is fed into the multilayer perceptron model for training. After adjusting and optimizing the functions of each layer, the trained multilayer perceptron model is obtained.

5. The method for monitoring water stress in citrus seedlings according to claim 4, characterized in that, The processing of data in the dataset specifically includes: Sort the values ​​in each data entry from largest to smallest; For similar data, data from a single day of similar experiments are randomly extracted and synthesized to obtain new data for that type of experiment, and dissimilar data are mixed in proportion. The original data is shifted up or down by a few small units to fill in the boundaries between different classes.

6. The method for monitoring water stress in citrus seedlings according to claim 3, characterized in that, The acquisition of data from the soil temperature and humidity sensor specifically includes: Set the query frames for each soil temperature and humidity sensor in sequence, initialize the dataframes for storing data from each soil temperature and humidity sensor, check if the serial port is open, send the query frames for each soil temperature and humidity sensor one by one, receive the returned information, transcode each one, calculate the temperature and humidity, obtain the current time, and write the time and temperature and humidity into the dataframes corresponding to each soil temperature and humidity sensor.

7. The method for monitoring water stress in citrus seedlings according to any one of claims 3-6, characterized in that, After feeding the first monitoring dataset into the trained multilayer perceptron model to obtain the evaluation results of seedling growth over a first preset time period, the process further includes: If the assessment result indicates an unhealthy state, and the assessment result is drought, it is recommended that the first pre-set irrigation in the future increase the soil moisture content within a healthy range to allow the plants to replenish and store moisture. If the assessment result is waterlogging, it is recommended that the first pre-set irrigation in the future decrease the soil moisture content within a healthy range to slow down the tendency of the plant roots to continue to grow elongate, increase the radial growth tendency of the plant roots, and make the plant roots grow stronger.

8. A computer-readable storage medium storing a program, characterized in that, When the program is executed by the processor, it implements the method for monitoring water stress in citrus seedlings as described in any one of claims 3-7.

Images