A stage irrigation control system based on plant factory seedling raising

Abstract

This application relates to the field of agricultural cultivation technology and discloses a phased irrigation control system based on plant factory seedling cultivation, comprising: a cultivation unit for providing cultivation space for crops; a spray unit for spraying the crops in the cultivation unit; a drip irrigation unit for drip irrigation the crops in the cultivation unit; a thermal imaging unit for acquiring thermal images of the cultivation unit; and a control unit for: instructing the thermal imaging unit to acquire subsequent thermal image time series of the cultivation unit; calculating the root activity index for the cultivation unit; and determining whether to switch the irrigation mode of the cultivation unit from spraying to drip irrigation. This invention quantifies the root activity of each seedling through the thermal imaging unit and makes independent asynchronous irrigation switching decisions accordingly. This ensures that each seedling switches from spraying to drip irrigation at its optimal physiological development point, solving the drawbacks of uniform switching and improving seedling uniformity and final quality.

Description

Technical Field

[0001] This invention relates to the field of agricultural cultivation technology, specifically to a phased irrigation control system based on plant factory seedling cultivation. Background Technology

[0002] Plant factories provide an ideal platform for the industrialized, standardized, and efficient cultivation of various crops (including but not limited to vegetables, flowers, trees, and rice) by precisely controlling environmental factors. During the seedling cultivation process, irrigation methods (including sprinkler and drip irrigation) are key factors affecting germination rate, seedling quality, and disease occurrence.

[0003] Sprinkler irrigation achieves large-area, uniform substrate surface wetting by spraying a mist or fine stream of water downwards from above the cultivation area. This characteristic gives it a significant advantage in the early germination stages after seed sowing, providing a consistent and sufficient moisture environment for all seeds and ensuring uniform emergence, which is especially important for small seeds or water-sensitive crop varieties. However, sprinkler irrigation also has significant inherent drawbacks: First, it keeps seedling leaves moist for extended periods, significantly increasing the risk of outbreaks of seedling diseases such as damping-off and seedling blight caused by fungi or bacteria; second, high evaporation rates from the leaves and substrate surface lead to water waste; and third, because water remains concentrated on the surface, it discourages seedling roots from penetrating and exploring deeper into the substrate, easily resulting in shallow, underdeveloped floating roots, which negatively impacts the later vigor of the plant.

[0004] Drip irrigation technology delivers water and nutrient solutions directly to the rhizosphere of each cultivation unit at a precisely controlled flow rate through specialized drippers or drip lines. Its advantages include: efficient use of water and nutrients, i.e., precise fertigation; and because it only wets the root substrate, the leaves and stems of seedlings remain dry, greatly reducing the probability of disease. However, drip irrigation also has its disadvantages due to its fixed-point water supply. In the early stages of sowing, seeds are widely distributed on the surface of the substrate, but the drip irrigation outlet is fixed. This may result in some seeds not receiving the initial moisture needed for germination, leading to uneven emergence and a high failure rate. This problem is particularly severe when seeds require a large amount of water for germination or when using lightweight substrates with poor water retention.

[0005] Therefore, a significant technical contradiction exists in existing technologies: a single irrigation method cannot simultaneously meet the two core and mutually exclusive technical goals of uniform initial emergence and robust seedlings with fewer diseases throughout the entire seedling cycle. Sprinkler irrigation excels at the former but is ineffective at the latter, while drip irrigation is the opposite. There is an urgent need in this field for a universal technical solution that can overcome these limitations and dynamically and intelligently adjust irrigation strategies according to the actual physiological needs of crops at different growth stages, thereby unifying the entire seedling process and maximizing seedling quality and efficiency. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a phased irrigation control system based on plant factory seedling cultivation. This system solves the problems of existing technologies, which, when using a single irrigation method, cannot simultaneously meet the contradictory needs of seed germination and robust seedling growth, and lack precise switching criteria based on the actual physiological state of individual seedlings when using different irrigation methods in combination, thus leading to a decline in overall seedling quality.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a phased irrigation control system based on plant factory seedling cultivation, comprising:

[0008] Cultivation unit, sprinkler unit, drip irrigation unit, thermal imaging unit, and control unit as the core;

[0009] The control unit is used to control the spraying unit to spray the crop during the first growth stage, and to initiate the diagnosis and switching process after triggering a signal during the response diagnosis stage.

[0010] One of the core technical solutions of this invention lies in the fact that the diagnostic and switching process achieves asynchronous switching between the physiological function diagnosis of individual seedlings and the irrigation mode. Specifically, for at least one cultivation unit, the control unit instructs the drip irrigation unit to apply diagnostic drip irrigation once, and instructs the thermal imaging unit to acquire the subsequent thermal image time series of the cultivation unit. The technical principle is that the diagnostic drip irrigation causes localized cooling of the cultivation unit surface, and the subsequent temperature recovery rate depends simultaneously on both surface evaporation and root water absorption. The more active the root water absorption, the faster the local water is absorbed, and the faster the temperature recovers to the environmental equilibrium temperature. This invention proposes to quantify this temperature decay rate as a root activity index to directly and quantitatively characterize the physiological function of individual seedlings.

[0011] In a specific implementation, to eliminate the interference of ambient temperature fluctuations, the root activity index is calculated as follows: First, the average temperature of the cultivation unit in each frame of the thermal image time series is extracted and compared with the average temperature of a preset dry substrate reference area in the same frame, thereby generating a decay curve of the relative temperature difference over time for the cultivation unit. Then, this decay curve is fitted to a first-order exponential decay model. The slope of the fitted line is calculated using a log-linear regression method, and the negative of this slope is taken as the root activity index. Its calculation formula can be expressed as:

[0012] ;

[0013] After obtaining the root activity index, the control unit compares it with a preset seedling activity threshold to determine whether to switch the irrigation mode of the cultivation unit from sprinkler to drip irrigation. Another key technical aspect of this invention is that the aforementioned diagnostic and switching decisions are executed independently and asynchronously for each cultivation unit. For seedlings whose root activity index reaches or exceeds the threshold, their irrigation mode is switched; for seedlings that do not reach the threshold, their irrigation mode is maintained and re-evaluated in subsequent diagnostic cycles.

[0014] To improve the accuracy and efficiency of diagnosis, this invention further proposes an adaptive probing strategy. For cultivation units whose root activity index is below a threshold in the current diagnostic cycle, the control unit dynamically adjusts and increases the diagnostic water volume applied by the drip irrigation unit in the next diagnostic cycle based on the difference between its activity index and the threshold. This adjustment logic aims to enhance the probing intensity for seedlings with weaker response signals to obtain temperature decay data with a higher signal-to-noise ratio. The diagnostic water drip irrigation volume to be applied in the next diagnostic cycle is... It can be calculated using the following formula:

[0015] ;

[0016] In the formula, For the next diagnostic cycle In the middle, for cultivation units The calculated drip irrigation volume of diagnostic water to be applied; The drip irrigation volume used as a baseline for diagnosis is the standard volume used for initial diagnosis or verification of units that have already met the standards. This is a preset, dimensionless positive gain coefficient used to adjust the sensitivity of the diagnostic water pulse volume to the root activity index deviation. The preset threshold for seedling vigor is defined in the same way as the threshold used in the root activity diagnosis module and subsequent switching decisions. For the current diagnostic cycle In the middle, for cultivation units The calculated root activity index.

[0017] In one embodiment, the system of the present invention further includes a seedling bed, a seedling growth LED light, a nutrient solution, and a water pump. The cultivation unit, the spraying unit, the drip irrigation unit, and the thermal imaging unit are all located inside the seedling bed. The cultivation unit is located in the middle of the seedling bed, and the spraying unit, the thermal imaging unit, and the seedling growth LED light are all located above the cultivation unit. The drip irrigation unit is located between the cultivation unit and the spraying unit.

[0018] The nutrient solution is connected to the water pump, and a filter is installed between the nutrient solution and the water pump. The output end of the water pump is connected to the spray unit and the drip irrigation unit through pipes. A solenoid valve for the spray unit is installed between the water pump and the spray unit. The drip irrigation unit is provided with multiple water outlets, and each water outlet is provided with a solenoid valve for the drip irrigation unit.

[0019] Furthermore, this invention proposes an irrigation feedforward correction mechanism that applies individual physiological data acquired during the diagnostic phase to subsequent long-term cultivation strategies. When the irrigation mode of a cultivation unit is first switched to drip irrigation, the control unit records its root activity index at that time as the unit's final target activity index. Based on this final target activity index, a preset baseline irrigation amount is corrected to generate a personalized drip irrigation amount specific to that cultivation unit for subsequent drip irrigation cultivation stages. The drip irrigation volume can be calculated using the following formula:

[0020] ;

[0021] In the formula, For cultivation unit 2, after correction The amount of drip irrigation; This is a preset baseline irrigation amount, a standard value applicable to the entire second growth stage (seedling stage); Let be a dimensionless irrigation regulation gain coefficient, where It can be set to a positive or negative value. When it is positive, an incentive-based cultivation strategy is implemented, where seedlings with stronger root activity receive more irrigation. When it is negative, a water-control and root-strengthening strategy is implemented, where seedlings with stronger root activity receive less irrigation to stimulate root growth. This allows the system to achieve not only individualized irrigation but also strategic precision cultivation.

[0022] In some implementations, the generation of the diagnostic phase trigger signal is based on at least one of two conditions: the system's cumulative runtime reaches a preset threshold, or the population emergence rate obtained through image processing reaches a preset threshold.

[0023] In some embodiments, during the first growth stage, the control unit can also dynamically calculate and adjust the spray interval between two spraying operations based on environmental parameters collected in real time from temperature sensors, humidity sensors, and light intensity sensors, using an evaporation rate prediction model, in order to maintain the optimal moisture content of the substrate surface under varying conditions.

[0024] This invention provides a phased irrigation control system based on plant factory seedling cultivation. It has the following beneficial effects:

[0025] 1. This invention applies diagnostic drip irrigation through a drip irrigation unit and collects subsequent thermal image time series using a thermal imaging unit. It calculates a root activity index for each cultivation unit to quantify its root water absorption capacity. Based on an independent comparison of this index with a preset seedling activity threshold, it performs asynchronous switching decisions, ensuring that each seedling transitions from spraying to drip irrigation at its optimal physiological development point. This effectively avoids the uneven development caused by a one-size-fits-all switching, allowing strong seedlings to obtain a suitable root growth environment and weak seedlings to continue to grow, thereby maximizing the quality and efficiency of seedling cultivation.

[0026] 2. For seedlings whose root activity index has not yet reached the standard, this invention uses an adaptive exploration strategy module to dynamically increase the volume of diagnostic water applied in subsequent diagnostic cycles. This enhances the thermal response signal of the slow-developing seedlings and obtains temperature decay data with a higher signal-to-noise ratio, thereby making the calculation of the root activity index more accurate and ensuring that the decision of the system of this invention will not be misjudged due to weak physiological signals.

[0027] 3. This invention records the final compliance activity index of each seedling when it switches to the target level through the irrigation feedforward correction module. Based on this, it generates a personalized drip irrigation amount that runs through the entire subsequent drip irrigation stage. By setting different irrigation adjustment gain coefficients, users can also flexibly implement different strategies such as incentive cultivation or water control to strengthen roots. This allows the system to not only achieve individualization but also precision cultivation, thereby maximizing the application value of diagnostic data and helping to cultivate high-quality seedlings that meet specific production goals. Attached Figure Description

[0028] Figure 1 This is a schematic diagram of the system structure of the present invention;

[0029] Figure 2 for Figure 1 Enlarged view of point A in the image;

[0030] Figure 3 This is a schematic diagram of the system architecture of the present invention;

[0031] Figure 4 This is a schematic diagram of the functional modules of the control unit of the present invention;

[0032] Figure 5 This is a schematic diagram of the adaptive exploration strategy of the present invention.

[0033] The components include: 1. Seedling bed; 2. Cultivation unit; 3. Sprinkler unit; 4. Drip irrigation unit; 5. Thermal imaging unit; 6. Seedling growth LED light; 7. Control unit; 8. Sprinkler unit solenoid valve; 9. Drip irrigation unit solenoid valve; 10. Nutrient solution; 11. Filter; 12. Water pump. Detailed Implementation

[0034] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0035] To better understand the present invention, the above content will be described in detail below with reference to specific embodiments.

[0036] Please see the appendix Figure 1 -Appendix Figure 5 This invention provides a phased irrigation control system based on plant factory seedling cultivation, the system comprising:

[0037] The cultivation unit 2 provides cultivation space for crops; the spraying unit 3 sprays water onto the crops in the cultivation unit 2; the drip irrigation unit 4 drip irrigates the crops in the cultivation unit 2; the thermal imaging unit 5 acquires thermal images of the cultivation unit 2; and the control unit 7.

[0038] In one embodiment, the system of the present invention further includes: a seedling bed 1, a seedling growth LED light 6, a nutrient solution 10, and a water pump 12. The cultivation unit 2, spray unit 3, drip irrigation unit 4, and thermal imaging unit 5 are all located inside the seedling bed 1. The cultivation unit 2 is located in the middle of the seedling bed 1, the spray unit 3, thermal imaging unit 5, and seedling growth LED light 6 are all located above the cultivation unit 2, and the drip irrigation unit 4 is located between the cultivation unit 2 and the spray unit 3. The nutrient solution 10 is connected to the water pump 12, and a filter 11 is installed between the nutrient solution 10 and the water pump 12. The output end of the water pump 12 is connected to the spray unit 3 and the drip irrigation unit 4 through pipes, and a spray unit solenoid valve 8 is installed between the water pump 12 and the spray unit 3. The drip irrigation unit 4 has multiple water outlets, and each water outlet is equipped with a drip irrigation unit solenoid valve 9. Specifically:

[0039] The seedling growth LED light 6 provides illumination for the crops in the cultivation unit 2. At the same time, the spray unit solenoid valve 8, the drip irrigation unit solenoid valve 9, the seedling growth LED light 6, the water pump 12, and the thermal imaging unit 5 are all connected to the control unit 7. The connection can be made by wired or wireless communication. Furthermore, the water outlet on the drip irrigation unit 4 is close to the roots of the plants. Therefore, by controlling the spray unit solenoid valve 8 or the drip irrigation unit solenoid valve 9 through the control unit 7, the spray unit 3 can be driven to perform spraying operations, or a certain water outlet on the drip irrigation unit 4 can be driven to perform drip irrigation operations. At the same time, the water pump 12 can be controlled to supply water to the drip irrigation unit 4 and the spray unit 3, and the seedling growth LED light 6 and the thermal imaging unit 5 can also be driven to perform operations.

[0040] In one embodiment, the control unit 7 is the core component and can be a PLC, an embedded system, or a central computer. It has pre-stored or can receive programs for different crop growth stages and can automatically control the start, stop, and switching of the sprinkler unit 3 and the drip irrigation unit 4 according to the time nodes set in the program or the signals fed back by sensors (such as humidity sensors or vision sensors).

[0041] Specifically, the control unit 7 may include a growth stage timing control module, which retrieves relevant growth model parameters from its internal database based on user-inputted crop type information or cultivation batch information, and determines the current macroscopic growth stage of the seedling process. In this invention, the growth stage timing control module divides the entire seedling cycle into two basic stages: the first growth stage, i.e., the germination stage; and the second growth stage, i.e., the seedling stage. This growth stage timing control module is mainly responsible for the irrigation mode management of the first growth stage and the transition triggering to the second growth stage.

[0042] During the first growth stage (germination period), the growth stage timing control module, through the operation of spray unit 3, sends control commands to the water pump and related valve group of spray unit 3 according to a preset program to execute periodic spraying operations. This program includes two core parameters: spraying duration. and spray interval These two parameters can be preset by the user or automatically loaded from the corresponding growth model parameters to maintain a uniform and suitable humidity environment on the entire surface of the seedbed 1, so as to promote uniform and rapid seed germination.

[0043] In one embodiment, when the spraying unit 3 performs the spraying step, it does not use a fixed, unchanging interval time. Specifically, when the growth stage timing control module performs the spraying management during the germination period, its real-time goal is to maintain a stable and optimal moisture level on the surface of the seedling substrate. This is to avoid seed germination failure due to dryness, and also to prevent seed rot or lack of oxygen due to excessive moisture.

[0044] To achieve this goal, the control unit 7 continuously collects real-time data from one or more environmental sensors, which may include temperature sensors, humidity sensors, and light intensity sensors deployed within the plant factory. Using these real-time environmental parameters as input, the control unit 7 dynamically calculates when the next spraying operation should be performed, based on a built-in evaporation rate prediction model.

[0045] The core of this dynamic adjustment lies in the spray interval. Real-time recalculation, and spray duration This interval is typically kept at a fixed value (e.g., 30 seconds) to ensure that each spray evenly wets the entire surface. The calculation logic for the spray interval is as follows:

[0046] First, a baseline spraying interval is preset in the growth model parameters associated with a specific crop. (e.g., 2 hours). This baseline value represents the spraying frequency required to maintain optimal moisture on the substrate surface under ideal, standard temperature, humidity, and light conditions (e.g., temperature 22°C, humidity 70%, light off).

[0047] Subsequently, the control unit 7 calculates a series of dimensionless adjustment coefficients based on the real-time collected environmental parameters to correct the reference interval:

[0048] Temperature regulation coefficient Ambient temperature is the most significant factor affecting the rate of water evaporation. Higher temperatures result in faster evaporation, requiring shorter spray intervals. This coefficient can be calculated using the following formula:

[0049] ;

[0050] In the formula, The preset reference temperature in the growth model. This is the current real-time temperature. When the real-time temperature is higher than the reference temperature, Less than 1, thus shortening the spray interval.

[0051] Humidity regulation coefficient The relative humidity of the environment directly affects the water vapor pressure difference in the air; the lower the humidity, the faster the evaporation. This coefficient can be calculated by the following formula:

[0052] ;

[0053] In the formula, The baseline relative humidity is This represents the current real-time relative humidity. When the real-time humidity is lower than the baseline, If the value is less than 1, the spray interval will also be shortened.

[0054] Light adjustment coefficient Light, especially light containing the infrared band, provides additional energy to accelerate water evaporation. This coefficient can be designed as a function related to light intensity. For example, a simplified model could be:

[0055] ;

[0056] In the formula, This is the current light intensity reading. This is a preset gain coefficient. As the light intensity increases, Reduce or shorten the spray interval.

[0057] Finally, the dynamically calculated spray interval It is derived from the following formula:

[0058] ;

[0059] Furthermore, regarding the substrate water-holding capacity parameter, since it is usually fixed during a single seedling stage, it is not used as an input for real-time adjustment, but rather as the initial basis for selecting or setting the baseline spraying interval. For example, when setting up the system, the user can select the substrate type (such as rock wool, peat moss, or sponge), and the system will retrieve the corresponding baseline spraying interval value from the database for the substrate with the appropriate water-holding capacity. For substrates with poor water-holding capacity (such as rock wool with larger particles), the baseline spraying interval will be set to be shorter.

[0060] Through the above mechanism, after completing one spray, the control unit 7 will accurately calculate the execution time of the next spray based on the comprehensive evaporation potential of the current environment, thus forming a responsive and precise closed-loop feedback system to ensure that the germinating seeds can always obtain the most stable and suitable water environment under changing environmental conditions.

[0061] Furthermore, the growth stage timing control module generates a diagnostic stage trigger signal upon determining the end of the first growth stage. This is used to activate the subsequent root activity diagnostic module. The triggering conditions are determined based on time or by collecting macroscopic population characteristics of the crop through a camera. When the conditions are met, the system considers the seedlings to have reached the preliminary state where they can enter the diagnostic switching stage.

[0062] This complex logic condition can be expressed by the following formula:

[0063] ;

[0064] In the formula, It is a Boolean variable; when its value is True, it means that the diagnostic phase has been triggered. This refers to the cumulative system runtime since the start of seeding; The maximum duration of the first growth stage (germination period) is determined based on time-based criteria. To obtain the population germination rate using conventional image processing methods, in one specific embodiment, this value can be obtained by converting the color space (e.g., from RGB to HSV) of a visible light image of the seedbed 1 using a camera, and then counting the proportion of green pixels to the total number of pixels in the cultivation area. This image processing method is well-known in the art and will not be described in detail here. The preset macro-population emergence rate threshold, such as 70%, is the judgment condition based on the macro-population characterization.

[0065] In one embodiment, the growth model parameters may specifically be a set of pre-defined or machine-constructed structured datasets associated with the standard growth process of a specific crop variety in a plant factory environment. This dataset is stored in the internal database of the control unit 7 and is retrieved and activated when the user selects a crop variety or cultivation batch.

[0066] The parameter set of this growth model mainly includes the following three types of parameters, which are used to guide the automated operation of the system:

[0067] 1. Baseline parameters for stage division and switching: These parameters provide a baseline for macroscopic regulation of the growth stage timing control module.

[0068] Maximum duration of the first growth stage (germination period) This is a time threshold derived from a large amount of historical planting data. For example, for a specific variety of lettuce, this value can be set to 72 hours. It can be used as one of the compound logic conditions triggered in the diagnostic stage to ensure that even if the germination rate test shows abnormalities, the system can force the system to enter the next stage after a reasonable time window, preventing problems caused by excessively long spraying time.

[0069] Macro-population emergence rate threshold This represents the percentage of the population that has reached a state where diagnostic switching can be initiated. For example, it can be set to 70%. Also determined based on crop characteristics and production experience, this ensures that more refined individual diagnostics are initiated only after most seedlings have emerged, thus improving system efficiency.

[0070] 2. Environmental control parameters for the first growth stage (germination period): These parameters provide specific operating instructions for the growth stage timing control module when executing the spraying task.

[0071] Spraying duration This refers to the duration of a single spraying operation, such as 30 seconds. This parameter is set to ensure that each spray completely wets the substrate surface without causing excessive moisture loss or waterlogging.

[0072] Spray interval This refers to the time interval between two spraying operations, such as 2 hours. This parameter is closely related to ambient temperature and humidity, light intensity, and the water-holding capacity of the substrate. Its purpose is to dynamically maintain the substrate surface in an optimally moist state for seed germination throughout the germination period.

[0073] 3. Physiological performance benchmark parameters in the diagnosis and drip irrigation stages: These parameters provide core physiological judgment and decision-making benchmarks for the root activity diagnosis module, adaptive exploration strategy module, and irrigation prescription feedforward correction module.

[0074] Seedling activity threshold This is one of the most critical physiological indicator thresholds in this invention. A quantitative standard for qualified root development is defined. This value can be constructed as follows: Under laboratory conditions, select a batch of standard-grown seedlings with confirmed good root development (e.g., through physical digging and observation), conduct an active heat pulse probing experiment on them, measure their root activity index, and take the statistical average or a certain percentile value (e.g., the 80th percentile) as the root activity index. For example, for lettuce seedlings, this value might be labeled as 0.025 (unit: s). -1 ).

[0075] Reference value of maximum root activity index : Normalization used in drip irrigation volume calculations. Represents the upper limit of root activity achievable by this crop variety under ideal conditions. It can be a theoretically derived value or the highest final target root activity index recorded in historical cultivation data.

[0076] Benchmark diagnostic water volume This defines the micro-dose infusion volume used for initial or standard diagnostics, for example, 0.5 ml. This value must be set in a balance between generating a sufficiently clear thermal signal and avoiding excessive interference with the substrate moisture state.

[0077] Standard drip irrigation volume It is used to define the standard amount of irrigation a standard seedling should receive per unit time (or per irrigation event) after entering the drip irrigation stage, for example, 2 ml per irrigation.

[0078] Therefore, by constructing the above growth model parameters, manufacturers can pre-set standard parameter packages for multiple common crops before the system leaves the factory; users can also continuously optimize and customize parameter sets for specific varieties based on their own production practices through the system's learning and calibration functions, thereby making the system highly adaptable and scalable.

[0079] When the growth stage timing control module determines that the crop meets the diagnostic stage triggering conditions and generates a diagnostic stage trigger signal. Afterwards, the control unit 7 will drive the root activity diagnosis module to collect crop production status through the thermal imaging unit 5. Specifically, in one collection process, the crop is first irrigated by the drip irrigation unit 4 to diagnose its status and determine whether it has reached the drip irrigation state (seedling stage). If it has, the cultivation unit 2 where the crop is located will be switched from spray to drip irrigation. If not, it will continue to be sprayed (and will be diagnosed by drip irrigation through the drip irrigation unit 4 at certain time intervals (e.g., once every 2 hours). Specifically, the drip irrigation unit 4 first drips a precise small amount of diagnostic water into each cultivation unit 2 to irrigate the crop, and then observes the data through a preset observation window. Inside, thermal imaging unit 5 is activated and continuously captures images at a fixed frame rate, with each frame of thermal image forming a data matrix. Therefore, the complete input data stream received by this root activity diagnostic module is a time-series collection. It can be represented as:

[0080] ;

[0081] In the formula, In the observation window The first Each sampling time The captured single-frame thermal image data matrix, where each element represents the temperature value at the corresponding physical location. This represents the total number of sampling frames within the observation window.

[0082] Subsequently, the root activity diagnostic module processes the aforementioned input data stream. After processing and calculation, a two-dimensional root activity index matrix is ​​finally generated. This is its output. The dimension of this output matrix corresponds to the layout of cultivation units 2 in seedbed 1, and each element in the matrix... That is, the coordinates defined in this invention for quantizing coordinates. Root activity index of seedling root water absorption capacity in cultivation unit 2 Its structure can be represented as:

[0083] ;

[0084] in, and These represent the number of rows and columns of a single-cell seedling tray, respectively, and this root activity index matrix... This constitutes a root activity map of the physiological state of all individual seedlings in the entire seedbed 1.

[0085] In one specific embodiment, the root activity diagnostic module processes the aforementioned input data stream. The processing and calculation specifically include:

[0086] The root activity diagnosis module receives the input data stream. Afterwards, the root activity diagnosis module first performs data preprocessing, specifically including image registration and region calibration. Image registration is used to correct pixel displacement caused by factors such as minor equipment vibrations between consecutive frames, ensuring accuracy throughout the observation window. Within the image, the same pixel always corresponds to the same physical location. Region calibration maps the pixel coordinates on the thermal image to the coordinates of the logical cultivation unit 2 in the seedling bed 1. This calibration process only needs to be completed once during the initial system installation or when the tray specifications are changed. It will determine a unique pixel area for each cultivation unit 2. The specific implementation of image registration and ROI extraction can be accomplished by those skilled in the art using well-known image processing libraries; this is a conventional technique in the field and will not be elaborated upon here.

[0087] After pretreatment, the root activity diagnostic module performs tests on each cultivation unit 2. The corresponding ROI, in each frame of the thermal image within the observation window. The average temperature value of all pixels within the unit is calculated to obtain the value representing the temperature of the cultivation unit 2 at time t. single temperature value And by analyzing the time series... By processing each frame image individually, a discrete temperature-time data curve can be generated for each cultivation unit 2.

[0088] To eliminate the interference of ambient temperature fluctuations on the measurement results, the root activity diagnostic module then needs to perform normalization processing on the acquired temperature curves. This processing is achieved by calculating the relative temperature difference between the surface of cultivation unit 2 and the surrounding dry, undampened substrate surface. This is achieved by treating the temperature of the drying substrate as the ambient thermal equilibrium temperature at the current moment. Therefore, for any cultivation unit 2 At any sampling time The relative temperature difference is calculated as follows:

[0089] ;

[0090] In the formula, At any moment The average temperature is extracted from a pre-defined reference area representing the dry substrate. Through this step, the original absolute temperature curve is transformed into a relative temperature difference curve reflecting the localized cooling effect caused by moisture evaporation and root absorption.

[0091] Furthermore, the core of this processing logic lies in performing kinetic analysis on the relative temperature difference curve to quantitatively calculate the root activity index. The physical basis for this lies in the relative temperature difference after drip irrigation is performed in drip irrigation unit 4. The decay process is mainly determined by water evaporation and root absorption. The more active the root system in water absorption, the faster the local water loss and the faster the temperature recovers to the environmental equilibrium temperature. The greater the decay rate, the better. This decay process can be approximated by a first-order exponential decay model. To solve for the rate constant of this model, i.e., the root activity index, the root activity diagnosis module uses a log-linear regression method. Taking the natural logarithm of the first-order exponential decay model, we obtain the linear relationship:

[0092] ;

[0093] In the formula, This is the initial observation time after the water pulse is applied;

[0094] And for each cultivation unit 2 In the observation window Internal collection Data points The root activity diagnosis module can use the least squares method to fit the above linear model, calculate the slope of the regression line, and then calculate the root activity index. This is the opposite of the slope. The formula for its calculation is as follows:

[0095] ;

[0096] Therefore, the root activity diagnostic module is used for all cultivation units 2 on seedbed 1. Repeat the entire calculation process described above, and eventually all the calculated root activity indices can be obtained. Compiled into the two-dimensional root activity index matrix mentioned above This is then used as the output of the root activity diagnostic module.

[0097] The root activity diagnosis module completes the calculation and outputs the root activity index matrix. Subsequently, the control unit 7 executes the asynchronous switching decision for this round based on the matrix. On the other hand, before the next diagnostic cycle is scheduled to be executed, the adaptive exploration strategy module is activated to dynamically optimize the execution parameters of the next diagnosis by analyzing the results of the current diagnosis, thereby forming a closed-loop adaptive control of the diagnostic behavior.

[0098] Specifically, the decision-making basis of this adaptive exploration strategy module is the root activity map generated by the root activity diagnosis module in the previous round, namely the root activity index matrix. Above. Specifically, for cultivation unit 2 whose irrigation mode has not yet been switched to drip irrigation before the start of the next diagnostic cycle, the adaptive exploration strategy module will read its most recently calculated root activity index. and compared it with the preset seedling activity threshold. A comparison is made. The result of this comparison directly determines the intensity of the probe used in the next diagnostic test of a particular cultivation unit 2.

[0099] Therefore, based on the above decision-making criteria, the core function of the adaptive detection strategy module is to generate and output a set of dynamically adjusted instructions. These instructions are used to set the physical parameters for the micro-dose drip irrigation performed by the drip irrigation unit 4 in the next diagnostic cycle. In this embodiment, the main parameter adjusted is the volume of water discharged by the drip irrigation unit 4. The adjustment logic is as follows: when the root activity index of a cultivation unit 2 is lower than the threshold, it means that when the thermal imaging unit 5 collects thermal images, the response signal of the water dripped from the drip irrigation unit 4 is weak. In order to obtain a temperature decay curve with a higher signal-to-noise ratio in the next diagnosis for more accurate calculation, the adaptive detection strategy module will instruct the drip irrigation unit 4 to apply a larger volume of diagnostic water to it.

[0100] Specifically, for the current diagnostic cycle Any cultivation unit 2 that has not yet met the switching criteria In the next diagnostic cycle The diagnostic water pulse volume to be used in the middle It can be calculated using the following formula:

[0101] ;

[0102] In the formula, For the next diagnostic cycle In the middle, regarding the first The volume of diagnostic water to be applied for drip irrigation calculated for each cultivation unit 2; The drip irrigation volume used as a baseline for diagnosis is the standard volume used for initial diagnosis or verification of units that have already met the standards. This is a preset, dimensionless positive gain coefficient used to adjust the sensitivity of the diagnostic water pulse volume to the root activity index deviation. The preset threshold for seedling vigor is defined in the same way as the threshold used in the root activity diagnosis module and subsequent switching decisions. For the current diagnostic cycle In the middle, regarding the first The root activity index calculated for cultivation unit 2.

[0103] After calculating the next round of diagnostic water volume for all cultivation units 2 that need adjustment, the adaptive detection strategy module converts these volume parameters into control signal parameters corresponding to each solenoid valve in the drip irrigation unit 4 (e.g., the duty cycle or on-time of the pulse width modulation (PWM) signal) and forms a set of output commands. At the start of the next diagnostic cycle, the control unit 7 will drive the drip irrigation unit 4 to apply the adjusted diagnostic water to different cultivation units 2 according to these commands, thereby achieving adaptive optimization of the diagnostic process itself.

[0104] Furthermore, while the adaptive exploration strategy module optimizes the diagnostic process, this invention further incorporates an irrigation feedforward correction module. This module maximizes the value of data acquired during the diagnostic phase that characterizes the individual physiological potential of seedlings, transforming it from a momentary judgment criterion for phase switching into a long-term, personalized cultivation guidance strategy that runs throughout the entire subsequent growth cycle.

[0105] Specifically, the irrigation feedforward correction module receives the root activity index recorded by the system for each cultivation unit 2 at the point of final achievement during the diagnostic switching phase. Specifically, when any cultivation unit 2... In a certain diagnostic cycle, the root activity index was calculated. The seedling viability threshold is reached or exceeded for the first time. At the same time, while performing the irrigation mode switch, the control unit 7 uses the root activity index value at this moment as the final target activity index for that unit. Permanent records are made. The recording process is ongoing until all cultivation units 2 on seedbed 1 have completed mode switching, thereby forming a complete dataset containing the physiological status of all seedlings at key developmental nodes.

[0106] Subsequently, based on the recorded data, the irrigation feedforward correction module will also generate output instructions to correct the baseline irrigation for each cultivation unit 2 throughout the entire drip irrigation phase. Specifically, a uniform baseline irrigation amount applicable to all plants will be generated. Adjusted to a personalized drip irrigation volume for each plant, matching its root development level. This correction process utilizes information obtained from early diagnosis to predict and adjust future control inputs in advance to achieve better breeding results.

[0107] Furthermore, the calculation of this personalized drip irrigation amount is achieved through a calibration algorithm. This algorithm compares the final target activity index of a cultivation unit 2 with its level when it just met the threshold; the greater the difference, the stronger its root development potential, and the larger the adjustment range of the irrigation amount. For any cultivation unit 2... Its drip irrigation volume It can be calculated using the following formula:

[0108] ;

[0109] In the formula, For cultivation unit 2, after correction The amount of drip irrigation; This is a preset baseline irrigation amount, a standard value applicable to the entire second growth stage (seedling stage); This is a dimensionless irrigation regulation gain coefficient, the magnitude of which determines the personalized cultivation strategy, specifically:

[0110] when At the same time, an incentive-based cultivation strategy was implemented: seedlings with stronger root activity received a larger amount of irrigation to support their faster growth.

[0111] when At the same time, a water-controlled root-strengthening strategy was implemented: the more active the seedlings, the less irrigation they received. By creating slight water stress, the roots were stimulated to explore deeper and wider, resulting in stronger root systems. Therefore, users can flexibly set irrigation levels according to their seedling cultivation goals. The value of enables this system to achieve not only individualized but also strategically precise cultivation.

[0112] Cultivation Unit 2 The final root activity index that meets the standard is recorded by the system when the irrigation mode is switched from sprinkler to drip irrigation; The preset threshold for seedling viability; This is a reference value for the maximum root activity index used for normalization. This value can be the theoretical maximum value preset according to the crop type, or the highest observed final target activity index that is dynamically updated in the current cultivation batch.

[0113] After all cultivation units 2 have completed the mode switch, the irrigation feedforward correction module will also... This is converted into a complete irrigation prescription matrix. Therefore, throughout the subsequent drip irrigation cultivation stage, the control unit 7 will strictly follow the values ​​in this matrix to issue commands to the solenoid valves of each drip irrigation unit 4.

Claims

1. A phased irrigation control system based on plant factory seedling cultivation, characterized in that, include: Cultivation unit (2) is used to provide cultivation space for crops; Spraying unit (3); used to spray the crops in cultivation unit (2); Drip irrigation unit (4) is used to drip irrigate the crops in cultivation unit (2); Thermal imaging unit (5); used to acquire thermal images of the cultivation unit (2); Control unit (7), which is used for: During the first growth stage, the control spray unit (3) sprays the crop and responds to the diagnostic stage trigger signal to start the diagnostic and switching process; In the diagnosis and switching process, for at least one cultivation unit (2), the drip irrigation unit (4) is instructed to apply a diagnostic drip irrigation, and the thermal imaging unit (5) is instructed to acquire the subsequent thermal image time series of the cultivation unit (2); Based on the thermal image time series, the root activity index is calculated for the cultivation unit (2); The root activity index is compared with the preset seedling activity threshold to determine whether to switch the irrigation mode of the cultivation unit (2) from spraying to drip irrigation. The control unit (7) includes: The growth stage timing control module is used to dynamically calculate and adjust the spraying interval between two spraying operations based on the environmental parameters collected in real time by the temperature sensor, humidity sensor and light intensity sensor in the first growth stage, through the evaporation rate prediction model. The control unit (7) further includes: a root activity diagnostic module, used to receive the thermal image time series collected by the thermal imaging unit (5) after the drip irrigation of diagnostic water; By extracting the average temperature of the cultivation unit (2) in each frame of the thermal image time series and comparing it with the average temperature of the preset dry substrate reference area in the same frame, a decay curve of the relative temperature difference over time is generated for the cultivation unit (2). The root activity index is calculated by processing the attenuation curve. The calculation of the root activity index includes: fitting the decay curve of the relative temperature difference over time to a first-order exponential decay model, calculating the slope of the fitted line using a log-linear regression method, and taking the negative of the slope as the root activity index. The control unit (7) further includes: performing independent asynchronous switching decisions on multiple cultivation units (2) on the seedbed (1), specifically: For any cultivation unit (2), when the root activity index reaches or exceeds the seedling activity threshold, its irrigation mode is switched from spraying to drip irrigation. If its root activity index is lower than the seedling activity threshold, its spraying status is maintained, and the diagnosis is performed again in the subsequent diagnostic cycle. The control unit (7) also includes an irrigation feedforward correction module, which records the root activity index at the time when the irrigation mode of a cultivation unit (2) is switched to drip irrigation for the first time as the final target activity index of the cultivation unit (2). And, based on the final qualified activity index, the drip irrigation amount for the subsequent drip irrigation cultivation stage is generated for the cultivation unit (2) by correcting the preset benchmark irrigation amount; The irrigation feedforward correction module specifically includes: Based on the amount by which the final qualified activity index exceeds the seedling activity threshold, and in conjunction with the reference value of the maximum root activity index, the normalized activity exceedance index is calculated. The normalized activity excess index is multiplied by the irrigation regulation gain coefficient to obtain a correction factor; Add the correction factor to 1 to form the correction coefficient; The drip irrigation amount is calculated by multiplying the baseline irrigation amount by the correction factor. The irrigation regulation gain coefficient can be set to a positive or negative value to increase or decrease the baseline irrigation amount, respectively.

2. The phased irrigation control system based on plant factory seedling cultivation according to claim 1, characterized in that, The growth stage timing control module further includes: generating the diagnostic stage trigger signal when the system's cumulative running time since sowing begins reaches the preset maximum duration of the first growth stage, or when the population emergence rate obtained through image processing reaches the preset macroscopic population emergence rate threshold.

3. The phased irrigation control system based on plant factory seedling cultivation according to claim 1, characterized in that, It also includes a seedling bed (1), a seedling growth LED light (6), a nutrient solution (10) and a water pump (12). The cultivation unit (2), the spray unit (3), the drip irrigation unit (4) and the thermal imaging unit (5) are all located inside the seedling bed (1). The cultivation unit (2) is located in the middle of the seedling bed (1). The spray unit (3), the thermal imaging unit (5) and the seedling growth LED light (6) are all located above the cultivation unit (2). The drip irrigation unit (4) is located between the cultivation unit (2) and the spray unit (3). The nutrient solution (10) is connected to the water pump (12), and a filter (11) is provided between the nutrient solution (10) and the water pump (12). The output end of the water pump (12) is connected to the spray unit (3) and the drip irrigation unit (4) through pipes respectively. A spray unit solenoid valve (8) is provided between the water pump (12) and the spray unit (3). The drip irrigation unit (4) is provided with multiple water outlets, and each water outlet is provided with a drip irrigation unit solenoid valve (9).

4. The phased irrigation control system based on plant factory seedling cultivation according to claim 1, characterized in that, The control unit (7) further includes an adaptive detection strategy module, which is used to dynamically adjust and increase the drip irrigation volume of diagnostic water applied by the drip irrigation unit (4) in the next diagnostic cycle for cultivation units (2) whose root activity index is lower than the seedling activity threshold in the current diagnostic cycle, based on the difference between the root activity index and the seedling activity threshold of the cultivation unit (2).

5. A phased irrigation control system based on plant factory seedling cultivation according to claim 4, characterized in that, The adaptive exploration strategy module specifically includes: Based on the relative difference between the current root activity index of the cultivation unit (2) and the seedling activity threshold, and combined with the preset positive gain coefficient, the adjustment term is calculated. Add the adjustment term to 1 to obtain the adjustment coefficient; The drip irrigation volume of the baseline diagnostic water is multiplied by the adjustment factor to calculate the drip irrigation volume of diagnostic water to be applied in the next diagnostic cycle.

Images