Dynamic optimization regulation and control method for grading environmental parameters of potato seedlings

Abstract

The invention discloses a potato seedling grading environment parameter dynamic optimization regulation and control method, and relates to the technical field of agricultural informationization. Based on an Internet of Things sensor deployed in a target area, stem diameter, chlorophyll content and plant height data of potato seedlings are collected, the seedlings are divided into weak seedlings, middle seedlings and strong seedlings by adopting K-means clustering, and the weak seedlings, the middle seedlings and the strong seedlings are classified into a weak seedling classification model, a middle seedling classification model and a strong seedling classification model; setting the duration of the initial light period; and carrying out gradient photoperiod test on each type of potato seedlings by taking the stem elongation rate and the chlorophyll content as constraint conditions. Through grading photoperiod modeling and ant colony algorithm dynamic optimization, the limitation of traditional fixed photoperiod regulation and control is broken through, the initial photoperiod range is set based on the seedling physiological difference, and the photoperiod-growth rate response curved surface model is constructed in combination with the stem elongation rate and chlorophyll content constraint conditions. The influence of different photoperiod combinations on seedling growth is quantified, accurate matching of photoperiods and seedling requirements is ensured, excessive growth or premature senility is avoided, and the growth rhythm stability and the resource utilization efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the field of agricultural information technology, and in particular to a method for dynamically optimizing and controlling environmental parameters of potato seedling grading. Background Art

[0002] Potatoes are a typical short-day crop, and their growth is regulated by the photoperiod. Short days can induce tuber differentiation, while long days can delay or inhibit tuber formation, causing the plants to grow too tall. Traditional open-air cultivation relies on seasonal light changes, resulting in a mismatch between the tuber formation period and market demand. Today, closed environments such as greenhouses and plant factories provide a technical basis for light regulation. Through supplementary light, shading or LED spectrum regulation, the limitations of natural light can be broken.

[0003] Traditional light duration control methods mostly use a fixed photoperiod, without considering the physiological differences of seedlings after grading, which can easily lead to excessive growth or premature aging of some seedlings due to photoperiod mismatch. In addition, the use of PID control has the problem of slow response speed, making it difficult to ensure the adaptability of the light in the seedling growth environment and the control scheme, which can easily cause rhythmic disorder of tuber formation. Therefore, how to analyze the seedling status, perform graded photoperiod modeling, and combine ant colony algorithm for ant colony path planning, analyze the promoting effect of each photoperiod combination on seedlings of different grades, and dynamically adjust the photoperiod duration, is the problem to be solved by the present invention. To this end, a dynamic optimization and control method for potato seedling graded environmental parameters is proposed. Summary of the Invention

[0004] The present invention aims to provide a method for dynamically optimizing and controlling environmental parameters of potato seedling grading to solve the problems raised in the above-mentioned background technology.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0006] A method for dynamically optimizing and controlling environmental parameters for potato seedling grading comprises the following steps:

[0007] Step 1: Based on IoT sensors deployed in the target area, data on stem diameter, chlorophyll content, and plant height of potato seedlings were collected. K-means clustering was used to classify the seedlings into three categories: weak seedlings, medium seedlings, and strong seedlings. The initial photoperiod duration was set (10-12 hours for weak seedlings, 12-14 hours for medium seedlings, and 14-16 hours for strong seedlings).

[0008] Step 2: Using stem elongation rate and chlorophyll content as constraints, conduct a gradient photoperiod test on each type of potato seedling, analyze the existing photoperiod combinations, and use multiple regression analysis to establish a photoperiod-growth rate response surface model to preliminarily determine the appropriate photoperiod interval;

[0009] Step 3: Based on the photoperiod-growth rate response surface model, the ant colony algorithm parameters and pheromone matrix are initialized, and the light intensity is introduced through the heuristic factor to optimize the path selection weight;

[0010] Step 4: Each ant in the ant colony algorithm selects a photoperiod path based on the pheromone concentration and the heuristic factor probability to screen the photoperiod combination;

[0011] Step 5: Using IoT sensors to collect data on stem diameter and chlorophyll content at fixed intervals, and analyzing the uniformity of seedling growth under each photoperiod path;

[0012] Step 6: Update pheromones every 24 hours based on the photoperiod path evaluation results, eliminate paths that lead to excessive growth or premature aging, and lock the optimal photoperiod-graded seedling matching scheme after 10 iterations to optimize and control the lighting parameters.

[0013] A further improvement of the technical solution of the present invention is that in step 1, the process of setting the duration of the initial light cycle is:

[0014] IoT sensors deployed in target areas simultaneously collected physiological data on potato seedlings, including stem diameter, chlorophyll content, and plant height. After data cleaning and standardization, a K-means clustering algorithm was used to classify the seedlings into three categories: weak, medium, and strong, based on the Euclidean distance between these physiological indicators.

[0015] According to the seedling grading results and the growth needs of potato seedlings, the initial photoperiod duration ranges are set for the three types of seedlings. Among them, the initial photoperiod duration ranges for weak seedlings, medium seedlings, and strong seedlings are 10-12h, 12-14h, and 14-16h, respectively. This ensures that seedlings in different growth states obtain differentiated lighting strategies and avoids uneven growth problems caused by unified regulation.

[0016] A further improvement of the technical solution of the present invention is that in step 2, the process of preliminarily determining the appropriate photoperiod interval is:

[0017] Using stem elongation rate (average daily growth) and chlorophyll content (SPAD value) as constraints, potato seedlings were grouped according to the previous grading results (weak seedlings, medium seedlings, and strong seedlings). A photoperiod gradient of 10-16 hours was set, with a step length of 2 hours. Each group was repeated three times and cultured continuously for 14 days under a controlled environment. Stem elongation rate (cm / d) and chlorophyll content (SPAD) were simultaneously monitored to ensure the temporal and spatial consistency of the data.

[0018] Based on the experimental data, photoperiod combinations that meet the constraints were screened. With photoperiod and seedling category (dummy variable coding) as independent variables, and stem elongation rate and chlorophyll content as dependent variables, a response surface model was fitted using multiple linear regression to construct a photoperiod-growth rate response surface model to quantify the nonlinear effect of photoperiod on growth indicators.

[0019] The model accuracy was evaluated by cross-validation (R 2 >0.85, RMSE <10%), verify the prediction accuracy of the model, output the nonlinear relationship between photoperiod and growth rate, and preliminarily determine the appropriate photoperiod interval.

[0020] A further improvement of the technical solution of the present invention is that the process of constructing the photoperiod-growth rate response surface model and preliminarily determining the appropriate photoperiod interval is as follows:

[0021] Experimental data on photoperiod and seedling category (categorical variables, coded as dummy variables: weak seedlings = 001, medium seedlings = 010, strong seedlings = 100), as well as the corresponding stem elongation rate and chlorophyll content, were simultaneously acquired. Outliers were removed to ensure data integrity. Continuous variables (photoperiod, stem elongation rate, chlorophyll content) were Z-score standardized to eliminate dimensional differences. Categorical variables (seedling category) were kept in dummy coding form and directly included. The acquired data were integrated to obtain a comprehensive dataset, which was divided into training and test sets in a 7:3 ratio.

[0022] With photoperiod and seedling category (dummy variable coding) as independent variables and stem elongation rate and chlorophyll content as dependent variables, a photoperiod-growth rate response surface model was constructed using multiple linear regression. The least squares method was used to fit the model, calculate the regression coefficient and its significance, and screen significant variables (p < 0.05).

[0023] Based on the regression coefficients, three-dimensional response surfaces of photoperiod-growth rate and photoperiod-chlorophyll content were generated, and the marginal effects of photoperiod on growth indicators including stem elongation rate and chlorophyll content were analyzed by partial derivatives. Among them, weak seedlings were more sensitive to photoperiod, and the chlorophyll content of strong seedlings decreased faster with the extension of photoperiod due to light inhibition. Segmented regression was introduced to deal with nonlinear relationships, and the photoperiod was divided into two segments: <12h and ≥12h. Models were fitted separately, and the AIC values ​​of the segmented model and the global model were compared. When ΔAIC>2, the segmented model was selected and determined as the optimal model. Then, the interval corresponding to the peak value ±0.5 standard deviation of the response surface was taken as the appropriate photoperiod interval for each seedling type.

[0024] A further improvement of the technical solution of the present invention is that in step 3, the process of optimizing the path selection weight is:

[0025] Based on the photoperiod-growth rate response surface model, the initial ant colony algorithm parameters were set, including the number of ants (a fixed number of individuals were assigned to each type of seedling), the pheromone volatility coefficient, and the initial pheromone concentration. A three-dimensional structure was constructed based on the seedling type (weak / medium / strong) and photoperiod combination to form a pheromone matrix, whose dimensions corresponded to the grid points of the photoperiod-growth rate response surface.

[0026] Light intensity is introduced into the ant colony algorithm as a heuristic factor, mapped to the interval [0,1] through normalization, and together with pheromone concentration, it constitutes the path selection probability formula, namely the transition probability. The heuristic factor weight is dynamically adjusted so that the guiding effect of light intensity on path selection decays with the number of iterations, avoiding local optimality while retaining the biological constraints of the photoperiod-growth rate response surface.

[0027] A further improvement of the technical solution of the present invention is that in step 4, the process of screening the photoperiod combination is:

[0028] Each ant locates the corresponding level of the pheromone matrix based on the current seedling type (weak / medium / strong seedlings). At the current location (photoperiod-seedling combination), it combines pheromone concentration and normalized light intensity to calculate the transfer probability of all feasible neighboring grid points. It then randomly selects the next location based on the transfer probability, balancing deterministic guidance and random exploration to avoid being stuck in a fixed path.

[0029] After the ants complete their path selection, they record the photoperiod-seedling combinations they have passed through, and after completing one iteration, they release pheromones along the path they have passed through.

[0030] Continuous iteration, each round updates the pheromone matrix and evaluates the path adaptability, sets a double termination condition, and after the iteration ends, extracts the grid point with the highest pheromone concentration in the corresponding photoperiod interval for each type of seedling from the final pheromone matrix as its optimal photoperiod matching scheme. Weak seedlings lock on long days to prevent premature aging, strong seedlings determine short days to promote tuber formation, and medium seedlings match the intermediate value for balanced growth.

[0031] A further improvement of the technical solution of the present invention is that the process of determining the optimal light cycle matching solution is:

[0032] The ant colony algorithm continuously iterates and updates the three-dimensional pheromone matrix. In each iteration, after the ant colony completes the path selection based on the current pheromone distribution and light heuristic information, it performs a global update on the three-dimensional pheromone matrix. The pheromone concentration of all grid points is attenuated according to the preset volatility coefficient to simulate the natural volatility process. The pheromone increment is superimposed on the corresponding grid points based on the growth rate prediction value of the path passed by the ants. At the same time, the path adaptability is re-evaluated after each iteration, and the comprehensive adaptability score is calculated. The growth rate prediction value of the photoperiod-seedling combination is associated with the pheromone concentration. This ensures that the matrix update is synchronized with the biological model, gradually guiding the search direction to converge to the high-fitness area, determining the optimal path, and improving the global search efficiency of the algorithm.

[0033] The algorithm sets dual termination conditions to balance computational efficiency and solution quality, namely, a preset maximum number of iterations and a dynamic convergence check. The preset maximum number of iterations prevents infinite loops. The dynamic convergence check determines that the algorithm has converged to a stable solution when the concentration value of the optimal path (the grid point with the highest pheromone concentration) does not change significantly for 10 consecutive iterations.

[0034] After the algorithm terminates, the optimal photoperiod matching scheme for each type of seedling is extracted from the final pheromone matrix. Weak seedlings are locked in the long-day interval, and apical dominance is suppressed and premature aging is prevented by extending the light. Strong seedlings determine the short-day interval, activate the expression of genes related to tuber formation, and promote the distribution of nutrients to underground organs. Medium seedlings match the intermediate photoperiod.

[0035] A further improvement of the technical solution of the present invention is that in step 5, the process of analyzing the uniformity of seedling growth under each photoperiod path is:

[0036] IoT sensors in the target area collect data on stem diameter and chlorophyll content at regular intervals of four hours, and simultaneously record corresponding photoperiod parameters. After uploading the data to the edge computing node, outliers are removed and missing values ​​are interpolated to generate a standardized time series dataset to ensure data integrity and consistency.

[0037] Based on the clustered seedling categories, they were grouped by photoperiod path. The coefficient of variation (CV) of stem diameter and the evenness index (calculated based on the entropy value of stem diameter data) of each seedling type under the same photoperiod path were calculated. The temporal dynamics of evenness were analyzed using a sliding window to identify the impact of photoperiod on growth stability. A graph of evenness changes over time under each path was generated.

[0038] Combined with the evenness analysis results, photoperiod paths with a stem diameter variation coefficient of <15% were screened as candidate options and marked as high-adaptability paths.

[0039] A further improvement of the technical solution of the present invention is that the process of analyzing the temporal dynamics of uniformity through a sliding window and generating a graph of the uniformity variation over time for each path is as follows:

[0040] Based on the seedling classification results obtained by K-means clustering, the data were grouped by photoperiod parameters to ensure that each group of seedlings was in the same photoperiod path. For each group of data, the coefficient of variation (CV) of stem diameter was calculated, which is the ratio of the standard deviation of stem diameter to the mean. This quantifies the degree of dispersion of stem diameter among seedlings under the same photoperiod. The smaller the CV value, the higher the growth consistency. At the same time, the evenness index was introduced to evaluate growth stability. The evenness index was calculated based on the entropy value of the stem diameter data.

[0041] The sliding window method was used to analyze the temporal trend of the evenness index. The window width was set to a fixed time interval (7 consecutive time points), and the window was slid point by point with a step size of 1. The mean of the evenness index within each window was calculated to generate time series data. By comparing the time series curves under different photoperiod paths, the variation pattern of evenness was identified. At the same time, the slope and fluctuation range of the evenness index under each path were calculated to quantify the difference in trend strength and stability.

[0042] The sliding window analysis results were visualized to generate a graph showing the change of the evenness index over time under each photoperiod path. The horizontal axis was time and the vertical axis was the evenness index value. Different paths were distinguished by different colors. By comparing the curves, the trend of the impact of the photoperiod on growth stability was presented.

[0043] A further improvement of the technical solution of the present invention is that in step 6, the optimization and control process of the lighting parameters is:

[0044] Every 24 hours, based on the photoperiod path evaluation results, the stem elongation rate and chlorophyll content of the seedlings under each photoperiod path were calculated. If the path resulted in a stem elongation rate > 0.5 cm / d (excessive growth) or a chlorophyll content < 30 SPAD (premature aging), it was marked as a poor-quality path and its pheromone concentration was reduced. Otherwise, it was marked as a high-quality path.

[0045] According to the daily evaluation results, the pheromone of the poor-quality paths is reduced at a fixed decay rate, and the pheromone of the high-quality paths is increased according to the fitness ratio. If a path is marked as poor-quality for three consecutive days, it is forced to be eliminated. If the pheromone ratio of the high-quality path exceeds the threshold of 80%, it is preferentially retained and its coverage is expanded. After 10 iterations, only the paths with the top 20% pheromone concentration are retained to form the candidate photoperiod set;

[0046] The candidate photoperiod set is grouped by seedling category, the photoperiod distribution of each group of paths is counted, the photoperiod range with the highest pheromone concentration is locked as the optimal matching solution, the optimal matching solution is input into the lighting control system, and the photoperiod is dynamically adjusted.

[0047] Due to the adoption of the above technical solution, the present invention has the following technical advancements compared to the prior art:

[0048] 1. The present invention provides a method for dynamic optimization and control of graded environmental parameters of potato seedlings. Through graded photoperiod modeling and dynamic optimization of ant colony algorithm, it breaks through the limitations of traditional fixed photoperiod control, sets the initial photoperiod range based on the physiological differences of seedlings, combines the stem elongation rate and chlorophyll content constraints, constructs a photoperiod-growth rate response surface model, accurately quantifies the effects of different photoperiod combinations on seedling growth, ensures that the photoperiod is accurately matched with seedling requirements, avoids excessive growth or premature aging, and improves the stability of growth rhythm and resource utilization efficiency.

[0049] 2. The present invention provides a method for dynamically optimizing and controlling environmental parameters of potato seedlings by graded levels. Through the photoperiod path optimization mechanism of the ant colony algorithm, the pheromone matrix is ​​used to record the promoting effect of each photoperiod combination on seedling growth, and the path selection weight is dynamically adjusted in combination with light intensity as a heuristic factor. A dynamic balanced search strategy is adopted to avoid local optimal solutions, ensuring that the algorithm efficiently converges to the global optimal photoperiod scheme within the iteration range. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0051] Figure 1 This is a schematic diagram of the workflow of a method for dynamically optimizing and controlling environmental parameters for potato seedling grading according to the present invention;

[0052] Figure 2 The present invention is a schematic diagram of a method flow for dynamically optimizing and controlling a method for grading environmental parameters of potato seedlings. DETAILED DESCRIPTION

[0053] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0054] Example 1, as Figure 1 、 Figure 2As shown, the present invention provides a method for dynamically optimizing and controlling environmental parameters of potato seedling grading, comprising the following steps:

[0055] Step 1: Based on the IoT sensors deployed in the target area, the stem diameter, chlorophyll content and plant height data of potato seedlings are collected. K-means clustering is used to divide the seedlings into three categories: weak seedlings, medium seedlings and strong seedlings. The initial photoperiod duration is set (10-12 hours for weak seedlings, 12-14 hours for medium seedlings, and 14-16 hours for strong seedlings). The physiological index data of potato seedlings including stem diameter, chlorophyll content and plant height are synchronously collected through the IoT sensors deployed in the target area. After data cleaning and standardization, The K-means clustering algorithm was used to classify seedlings based on the Euclidean distance of physiological indicators, dividing them into three categories: weak seedlings, medium seedlings, and strong seedlings. Based on the seedling classification results and the comprehensive growth requirements of potato seedlings, the initial photoperiod duration ranges for the three categories of seedlings were set. Among them, the initial photoperiod duration ranges for weak seedlings, medium seedlings, and strong seedlings were 10-12 hours, 12-14 hours, and 14-16 hours, respectively. This ensured that seedlings in different growth states received differentiated lighting strategies and avoided the problem of uneven growth caused by unified regulation.

[0056] The specific work content is: through the deployment of a high-precision IoT sensor network in the target planting area, key physiological indicator data of potato seedlings are synchronously collected, including stem diameter (reflecting the plant's nutritional reserves and mechanical support capacity), chlorophyll content (characterizing photosynthesis efficiency and health status) and plant height (reflecting vertical growth rate and population density). The collected data is wirelessly transmitted to the edge computing node for data cleaning, and outliers caused by sensor failure or environmental interference are eliminated. The Z-score standardization method is used to normalize the multi-dimensional data to eliminate dimensional differences. The standardized data is clustered. The analysis provides a unified input to avoid classification bias caused by different indicator magnitudes. In the target planting area (i.e., potato experimental field), IoT sensors are arranged in a 5m×5m grid. Stem diameter is monitored using a laser ranging sensor (accuracy ±0.01mm), which is installed 3cm from the base of the plant. The stem diameter is calculated by the reflection time difference to avoid mechanical contact damage. Chlorophyll content is monitored using a multispectral SPAD instrument to non-destructively measure the chlorophyll content of leaves. The third fully expanded leaf at the top of each plant is selected. Plant height monitoring deploys a ToF lidar to vertically scan the plant canopy and extract the plant height using a point cloud processing algorithm. The frequency was 1 Hz. The pre-processed stem diameter, chlorophyll content and plant height data were used as feature vectors. The K-means clustering algorithm was used for unsupervised classification based on Euclidean distance. The seedlings were divided into three categories: weak seedlings, medium seedlings and strong seedlings, which corresponded to low growth potential, medium growth potential and high growth potential groups respectively through iterative optimization of the intra-cluster distance minimization and inter-cluster distance maximization goals. The clustering quality of the classification results was evaluated by the silhouette coefficient to ensure that the intra-class compactness and inter-class separation were in line with expectations. After grading, the mean and standard deviation of the physiological characteristics of each type of seedling were recorded as feature labels. According to the seedling grading results, combined with the light cycle requirements for potato growth, the seedlings were classified into three categories: weak seedlings, medium seedlings and strong seedlings. To find out the characteristics, differentiated initial photoperiod ranges are formulated for different types of seedlings. Among them, weak seedlings have weak growth potential, so they need to extend the light time to enhance the accumulation of photosynthetic products. The initial photoperiod is set at 10-12 hours. Middle seedlings are in the transition stage from vegetative growth to reproductive growth. A medium photoperiod of 12-14 hours is used to balance the development of aboveground and underground parts. Strong seedlings have strong growth potential, and too long light may inhibit tuber formation. A short day range of 14-16 hours is set to induce tuber differentiation. By setting the photoperiod in a graded manner, problems such as premature aging of weak seedlings and excessive growth of strong seedlings under unified regulation are avoided, and an initial benchmark is provided for subsequent dynamic optimization.

[0057] Step 2, with stem elongation rate and chlorophyll content as constraints, a gradient photoperiod (decreasing every 2h) test was carried out on each type of potato seedlings, the existing photoperiod combinations were analyzed, and a photoperiod-growth rate response surface model was established using multiple regression analysis. The appropriate photoperiod interval was preliminarily determined, and the stem elongation rate (average daily growth) and chlorophyll content (SPAD value) were used as constraints. The potato seedlings were grouped according to the early classification results (weak seedlings / middle seedlings / strong seedlings), and a photoperiod gradient of 10-16h was set respectively, with a step length of 2h. Each group was repeated 3 times, and cultured continuously for 14 days under a controlled environment. The stem elongation rate (cm / d) and chlorophyll content (SPAD) were monitored synchronously to ensure the temporal and spatial consistency of the data. The controlled environment was a temperature of 25℃ / 15℃ day and night, a humidity of 60%, and a stem elongation rate of 10-16h. The growth rate was calculated by the difference of plant height measurements for three consecutive days; the chlorophyll content was measured using the SPAD value at a fixed time every day. Based on the experimental data, photoperiod combinations that met the constraints were screened, among which, the weak seedlings retained a stem elongation rate > 0.5 cm / d and a chlorophyll > 30 SPAD combination; the medium seedlings retained a rate of 0.3-0.5 cm / d and a chlorophyll of 25-30 SPAD combination; and the strong seedlings retained a rate of < 0.3 cm / d and a chlorophyll < 25 SPAD combination. With photoperiod and seedling category (dummy variable coding) as independent variables, and stem elongation rate and chlorophyll content as dependent variables, a response surface model was fitted by multiple linear regression to construct a photoperiod-growth rate response surface model to quantify the nonlinear effect of photoperiod on growth indicators. The model accuracy was evaluated by cross-validation (R 2 >0.85, RMSE <10%), verify the prediction accuracy of the model, output the nonlinear relationship between photoperiod and growth rate, and preliminarily determine the appropriate photoperiod interval;

[0058] The process of constructing the photoperiod-growth rate response surface model and preliminarily determining the appropriate photoperiod interval is as follows:

[0059] The experimental data of photoperiod, seedling category (categorical variables, coded as dummy variables: weak seedlings = 001, medium seedlings = 010, strong seedlings = 100), and the corresponding stem elongation rate and chlorophyll content were obtained simultaneously. Outliers were removed to ensure data integrity. Continuous variables (photoperiod, stem elongation rate, chlorophyll content) were Z-score standardized to eliminate dimensional differences. Categorical variables (seedling category) remained in dummy coding form and were directly included for use. The acquired data were integrated to obtain a comprehensive data set, which was divided into training and test sets in a 7:3 ratio. Photoperiod and seedling category (dummy variable coding) were used as independent variables, and stem elongation rate and chlorophyll content were used as dependent variables. Multiple linear regression was used to construct a photoperiod-growth rate response surface model. The least squares method was used to fit the model, and the regression coefficients and their significance were calculated. , screened significant variables (p<0.05), generated three-dimensional response surfaces of photoperiod-growth rate and photoperiod-chlorophyll content based on the regression coefficients, and analyzed the marginal effects of photoperiod on growth indicators including stem elongation rate and chlorophyll content through partial derivatives. Among them, weak seedlings were more sensitive to photoperiod, and the chlorophyll content of strong seedlings decreased faster with the extension of photoperiod due to light inhibition. Segmented regression was introduced to deal with nonlinear relationships, and the photoperiod was divided into two segments: <12h and ≥12h. Models were fitted separately, and the AIC values ​​of the segmented model and the global model were compared. If the AIC of the segmented model was lower than that of the global model (ΔAIC>2), the segmented model was considered more reasonable, that is, when ΔAIC>2, the segmented model was selected and determined as the optimal model. Then, the interval corresponding to the peak value ±0.5 standard deviation of the response surface was taken as the appropriate photoperiod interval for each seedling type.

[0060] The photoperiod-growth rate response surface model is divided into a stem elongation rate model and a chlorophyll content model, and the expression is as follows:

[0061] SER=β0+β1·T+β2·W+β3·M+β4·(T×W)+β5·(T×M)+∈;

[0062] SPAD=γ0+γ1·T+γ2·W+γ3·M+γ4·(T×W)+γ5·(T×M)+∈;

[0063] Where SER is the stem elongation rate, SPAD is the chlorophyll content, T is the photoperiod, W is the dummy variable for weak seedlings, M is the dummy variable for medium seedlings, β0 and γ0 are intercept terms (theoretical values ​​of the strong seedling baseline group when photoperiod = 0), β1 and γ1 are the marginal effects of photoperiod T on strong seedlings (baseline effects), β2 and γ2 are the intercept differences between weak seedlings and strong seedlings (offsets when photoperiod = 0), β3 and γ3 are the intercept differences between medium seedlings and strong seedlings, β4 and γ4 are the interaction effects of weak seedlings and photoperiod (T×W), β5 and γ5 are the interaction effects of medium seedlings and photoperiod (T×M), and ∈ is a random error term. In the global multiple linear regression model, based on the avoidance strategy of the dummy variable trap in statistics, the strong seedling dummy variable was not explicitly introduced. The baseline state of strong seedlings was represented by all dummy variables being 0, that is, W = 0 and M = 0. Strong seedlings were used as the reference category, and their effects were included in the intercept term β0 or γ0.

[0064] The marginal effect of stem elongation rate is expressed as follows:

[0065]

[0066] The marginal effect of chlorophyll content is expressed as follows:

[0067]

[0068] Explanation: For weak seedlings, for every 1-hour increase in photoperiod, the stem elongation rate changes by β1+β4, and the chlorophyll change by γ1+γ4. If β4>0, the weak seedlings are more sensitive to photoperiod (lean growth risk 1). For strong seedlings, for every 1-hour increase in photoperiod, the stem elongation rate changes by β1 (β1<0, inhibiting elongation).

[0069] The specific work involved grouping seedlings according to the preliminary classification results (weak seedlings, medium seedlings, and strong seedlings) using stem elongation rate and chlorophyll content as constraints. Each group was set to a photoperiod gradient (10, 12, 14, and 16 hours, with a 2-hour step) for three replicates. The seedlings were cultured for 14 days in a controlled environment with a daytime temperature of 25°C / 15°C and a humidity of 60%. Edge computing nodes regularly recorded the photoperiod settings (accurate to 0.1 hour) and seedling category labels. Stem elongation rate was calculated by the difference between plant height measurements over three consecutive days, and chlorophyll content was measured using the SPAD value at a fixed time each day. Based on the biological threshold of potato photoperiod requirements, photoperiod combinations that met the constraints were selected. For weak seedlings, combinations with a stem elongation rate >0.5 cm / day and a chlorophyll content >30 SPAD were retained. For medium seedlings, combinations with a stem elongation rate of 0.3-0.5 cm / day and a chlorophyll content of 25-30 SPAD were retained. For strong seedlings, combinations with a stem elongation rate <0. 3cm / d and chlorophyll <25SPAD, ensure that the seedling category is coded as a three-variable dummy variable, that is, weak seedlings = 001, medium seedlings = 010, and strong seedlings = 100, and Z-score standardization is performed on the continuous variables including photoperiod, stem elongation rate, and chlorophyll content. The categorical variable (seedling category) remains in the dummy coding form (001 / 010 / 100) and does not participate in standardization. The training set and test set are divided into a 7:3 ratio to ensure that each type of seedling is evenly distributed in the subset. At the same time, integrity check: ensure that there are no missing values ​​in the training set and test set (mean interpolation is used when the missing rate is <1%); a response surface model is constructed using multiple linear regression, with photoperiod and seedling category as independent variables and stem elongation rate and chlorophyll content as dependent variables. The nonlinear relationship between photoperiod and growth indicators is quantified, and segmented regression (photoperiod <12h and ≥12h) is introduced. The AIC value is compared to select the optimal model. The cross-validation requirement is R 2 >0.85, RMSE<10%, the optimal photoperiod interval was determined by the peak value ±0.5 standard deviation of the response surface, and the optimal photoperiod intervals for weak seedlings, medium seedlings, and strong seedlings were clarified;

[0070] Step 3: Based on the photoperiod-growth rate response surface model, the ant colony algorithm parameters and pheromone matrix are initialized, and light intensity is introduced as a heuristic factor to optimize the path selection weight. Based on the photoperiod-growth rate response surface model, the initialized ant colony algorithm parameters are set, including the number of ants (a fixed number of individuals are assigned to each type of seedling), the pheromone volatility coefficient, and the initial pheromone concentration. A three-dimensional structure is constructed according to the seedling type (weak / medium / strong) and the photoperiod combination to form a pheromone matrix, whose dimensions correspond to the grid points of the photoperiod-growth rate response surface. Light intensity is introduced into the ant colony algorithm as a heuristic factor and mapped to the [0,1] interval through normalization. Together with the pheromone concentration, it constitutes the path selection probability formula, i.e., the transition probability. The heuristic factor weight is dynamically adjusted so that the guiding effect of light intensity on path selection decays with the increase in the number of iterations, avoiding local optimality while retaining the biological constraints of the photoperiod-growth rate response surface.

[0071] The specific work content is as follows: Based on the photoperiod-growth rate response surface model, the core parameters of the ant colony algorithm including the number of ants, pheromone volatility coefficient and initial pheromone concentration are initialized. For different seedling categories (weak seedlings, medium seedlings, and strong seedlings), the number of ants is allocated in a fixed ratio to ensure that the search process of each type of seedling is independent and the resources are balanced. The pheromone volatility coefficient is set to a dynamic attenuation mode with a higher initial value to accelerate early exploration, and then gradually decreases with the increase in the number of iterations to balance the global search and local convergence capabilities. The initial pheromone concentration is set according to the predicted value of the response surface model. A higher initial concentration is given to the high growth rate area to guide the ants to explore the potential optimal solution first. When constructing a three-dimensional pheromone matrix, the photoperiod is used as the horizontal axis, the seedling category is used as the vertical axis, and the growth rate is used as the vertical axis. The response surface grid points are mapped to matrix elements. Each element stores the pheromone concentration of the corresponding photoperiod-seedling combination, forming a structured data storage framework consistent with the model dimension. The light intensity is introduced as a heuristic factor into the path selection mechanism. Light intensity data was normalized and mapped to the interval [0, 1] via linear transformation to eliminate dimensional differences and unify the numerical range. Pheromone concentration and normalized light intensity were combined in a weighted sum to form the path selection probability. The weight coefficient was dynamically adjusted with the number of iterations. Initially, light intensity was given a higher weight (0.7) to strengthen its guiding effect on ant paths. At this time, ants preferentially moved to high-light areas and quickly located potential high-growth rate areas. As the iterations progressed, the light weight decreased by 0.05 per round and stabilized at 0.3 (the pheromone weight increased to 0.7) in the later stages to prevent the algorithm from falling into a local optimum due to over-reliance on light intensity. At the same time, based on the biological characteristics of potato, the effective range of photoperiods for different seedling types was set. When ants moved, they were only allowed to select grid points within this range, eliminating invalid solutions that violated physiological characteristics. After each iteration, the growth rate prediction value for each photoperiod-seedling combination was recalculated based on the latest response surface model, and the pheromone matrix was updated simultaneously to ensure that the algorithm always conformed to the actual growth pattern.

[0072] Step 4: Each ant in the ant colony algorithm selects a photoperiod path based on the pheromone concentration and the heuristic factor probability to screen the photoperiod combination. Among them, weak seedlings give priority to long-day (14-16h) to prevent premature aging, strong seedlings focus on short-day (10-12h) to induce tubers, and medium seedlings use the intermediate value (12-14h) for balanced growth. Each ant locates the corresponding level of the pheromone matrix according to the current seedling category (weak / medium / strong seedlings), and calculates the transfer probability of all feasible neighbor grid points at the current position (photoperiod-seedling combination) by combining the pheromone concentration and the normalized light intensity. Weak seedlings first calculate the probability of the 14-16h long-day path, strong seedlings focus on the 10-12h short-day path, and medium seedlings evaluate the middle range of 12-14h. The probability weight is dynamically adjusted, with the early light intensity dominant and the later pheromone intensity dominant. The pheromone concentration is dominant, and the next position is randomly selected according to the transfer probability, taking into account both deterministic guidance and random exploration to avoid being trapped in a fixed path. After the ant completes the path selection, the light cycle-seedling combination it has passed is recorded. After the ant completes one iteration, the pheromone is released on the path it has passed. The increment value is positively correlated with the growth rate prediction value corresponding to the path. The high growth rate path obtains more information pheromone increments, and the inefficient path is weakened. Continuous iteration, each round updates the pheromone matrix and evaluates the path adaptability, sets a double termination condition, and after the iteration terminates, extracts the grid point with the highest pheromone concentration in the light cycle interval corresponding to each type of seedling from the final pheromone matrix as its optimal light cycle matching scheme. Weak seedlings lock on long days to prevent premature aging, strong seedlings determine short days to promote tubers, and medium seedlings match the intermediate value for balanced growth.

[0073] The process of determining the optimal photoperiod matching scheme is as follows:

[0074] The ant colony algorithm updates the three-dimensional pheromone matrix through continuous iteration. In each round of iteration, after the ant colony completes the path selection based on the current pheromone distribution and light heuristic information, the three-dimensional pheromone matrix is ​​globally updated. The pheromone concentration of all grid points is attenuated according to the preset volatility coefficient to simulate the natural volatility process. According to the predicted growth rate of the path passed by the ants, the pheromone increment is superimposed on the corresponding grid point. The high growth rate path obtains a higher increment, forming a cumulative mechanism of strengthening high-quality paths and eliminating inefficient paths. At the same time, the path adaptability is re-evaluated after each round of iteration, and the comprehensive adaptability score is calculated. The growth rate prediction value of the photoperiod-seedling combination is associated with the pheromone concentration to ensure that the matrix update is synchronized with the biological model, and the search direction is gradually guided to converge to the high fitness area, the optimal path is determined, and the global search efficiency of the algorithm is improved. The algorithm sets a dual termination condition to balance the computational efficiency and the quality of the solution, that is, the preset maximum number of iterations and dynamic convergence judgment. The preset maximum number of iterations is used to prevent infinite Cycle; through dynamic convergence judgment, when the concentration value of the optimal path (grid point with the highest pheromone concentration) has no significant change for 10 consecutive iterations, the algorithm is considered to have converged to a stable solution. During the termination judgment process, the pheromone distribution dynamics in the corresponding photoperiod interval of each type of seedling is continuously monitored. By comparing the stability of historical iteration data with the current optimal solution, the reliability of the termination decision is ensured, effectively avoiding suboptimal solution locking caused by premature termination or waste of computing resources caused by excessive iteration. After the algorithm terminates, the optimal photoperiod matching scheme for each type of seedling is extracted from the final pheromone matrix. Weak seedlings are locked in the long-day interval. By extending the light, apical dominance is suppressed and premature aging is prevented. Strong seedlings determine the short-day interval, activate the expression of genes related to tuber formation, and promote the distribution of nutrients to underground organs. Medium seedlings match the intermediate photoperiod to balance the needs of vegetative growth and reproductive growth. When extracting, the seedling-specific photoperiod requirement threshold is followed, and only the grid points with the highest pheromone concentration and in line with biological constraints are selected to ensure that the scheme has both mathematical optimality and physiological rationality.

[0075] The expression of the comprehensive adaptability score is as follows:

[0076] F op =α·τ op +β·η op ;

[0077]

[0078] Where, F op is the comprehensive adaptability score of the current grid point (photoperiod o-seedling category p combination), the higher the value, the better the path, τ op is the pheromone concentration of the grid point (o, p) in the pheromone matrix, reflecting the cumulative preference of the ant colony for the path in the historical iteration, and is dynamically adjusted by the pheromone update rule. opis the light heuristic information, which is defined as the matching degree between the normalized light intensity and the seedling requirements, L o is the duration of the current light cycle o, L p is the optimal photoperiod center value of seedling category p, α and β are weight coefficients, which control the contribution ratio of pheromone and light heuristic information;

[0079] The specific work content is: each ant is located at the corresponding level of the three-dimensional pheromone matrix according to the seedling category (weak seedling, medium seedling, strong seedling). The matrix uses the light period as the horizontal axis, the seedling category as the vertical axis, and the growth rate as the vertical axis to store the pheromone concentration of each combination. The ant evaluates the transfer probability of all feasible neighboring grid points at the current position. The calculation combines the pheromone concentration and the normalized light intensity, and the weight is dynamically adjusted. In the initial iteration stage, the light intensity weight accounts for a higher proportion (0.7), guiding the ants to prioritize exploring the light-suitable area. Later, as the number of iterations increases, the ant's weight is adjusted. , the weight gradually tilts towards the pheromone concentration (finally accounting for 0.7), strengthens the cumulative preference for high-quality paths, focuses on the probability calculation of the 14-16h long-day path for weak seedlings, focuses on the 10-12h short-day path for strong seedlings, and evaluates the 12-14h intermediate range for medium seedlings. The next position is randomly selected by probability, balancing deterministic guidance and random exploration to avoid falling into the local optimal solution; after the ants complete the path selection, the light cycle-seedling combination sequence they have passed is recorded. After each iteration, all ants release pheromones on their path, and the incremental value is proportional to the light cycle corresponding to the path. The predicted growth rate of each seedling combination is positively correlated, meaning that high-growth-rate paths receive higher pheromone increments, while inefficient paths are weakened due to volatility. The pheromone matrix is ​​dynamically updated with each iteration, and the global pheromone decays according to a fixed volatility coefficient. At the same time, the newly added pheromones of this iteration are superimposed, forming an "accumulation-elimination" mechanism. By strengthening high-value paths and weakening inefficient paths, the ant colony gradually converges to the potential optimal solution. After each iteration, the adaptability of each path is re-evaluated to ensure that the pheromone distribution remains synchronized with the growth rate prediction model and always reflects the latest search results. The ant colony algorithm continues to iterate until the termination condition is met, the preset maximum number of iterations is reached, or the pheromone concentration of the optimal path does not change significantly for 10 consecutive times. After termination, the grid point with the highest concentration within the corresponding photoperiod interval for each seedling type is extracted from the final pheromone matrix as the optimal solution. Among them, weak seedlings are locked to the long-day regimen to inhibit premature aging by extending the light intensity. Strong seedlings are determined to adopt the short-day regimen to promote tuber enlargement. Medium seedlings match the intermediate value to balance vegetative growth and reproductive growth. The photoperiod requirement thresholds of different seedlings are followed to ensure biological rationality.

[0080] Step 5: Using IoT sensors to collect data on stem diameter and chlorophyll content at a fixed time interval (every 4 hours) to analyze the uniformity of seedling growth under each photoperiod path;

[0081] Step 6: Update the pheromone every 24 hours based on the photoperiod path evaluation results, eliminate the paths that lead to excessive growth (stem elongation rate > 0.5 cm / d) or premature aging (chlorophyll content < 30 SPAD), and lock the optimal photoperiod-graded seedling matching scheme after 10 iterations to optimize the lighting parameters.

[0082] Example 2, as Figure 1 、 Figure 2 As shown, based on Example 1, the present invention provides a technical solution: Preferably, in step 5, the process of analyzing the uniformity of seedling growth under each photoperiod path is:

[0083] IoT sensors in the target area collect data on stem diameter and chlorophyll content at fixed intervals of every four hours, and the corresponding photoperiod parameters are recorded simultaneously. After the data is uploaded to the edge computing node, outliers are removed and missing values ​​are interpolated to generate a standardized time series dataset to ensure data integrity and consistency. Based on the clustered seedling categories, they are grouped by photoperiod path, and the coefficient of variation (CV) and uniformity index (calculated based on the entropy value of stem diameter data) of each type of seedling under the same photoperiod path are calculated. The temporal dynamics of uniformity are analyzed through a sliding window to identify the impact trend of photoperiod on growth stability. A curve chart of the change of uniformity over time under each path is generated. Combined with the uniformity analysis results, photoperiod paths with a stem diameter coefficient of variation <15% are screened as candidate solutions and marked as highly adaptable paths.

[0084] The process of analyzing the temporal dynamics of uniformity through a sliding window and generating a graph of the uniformity variation over time for each path is as follows:

[0085] Based on the seedling classification results obtained by K-means clustering, the data were grouped according to the photoperiod parameters to ensure that each group of seedlings was in the same photoperiod path. For each group of data, the coefficient of variation (CV) of stem thickness was calculated, that is, the ratio of the standard deviation of stem thickness to the mean, to quantify the degree of dispersion of stem thickness among seedlings under the same photoperiod. The smaller the CV value, the higher the growth consistency. At the same time, the uniformity index was introduced to evaluate the growth stability. The uniformity index was calculated based on the entropy value of the stem thickness data. Among them, by constructing a stem thickness frequency distribution histogram, the stem thickness data was divided into 10 equal-width intervals, and the probability distribution of each interval was calculated. The entropy value (i.e., uniformity index) was calculated using the Shannon entropy formula. The lower the entropy value, the more concentrated the data distribution and the higher the uniformity. The coefficient of variation of stem thickness and the uniformity index respectively characterized the growth stability from the perspectives of dispersion and distribution concentration. The sliding window method was used to analyze the temporal trend of the uniformity index. The window width was set to a fixed time interval (7 consecutive time points), and the window was slid point by point with a step size of 1 to calculate the value of each window. The mean of the intraoral uniformity index is used to generate time series data. By comparing the time series curves under different photoperiod paths, the change pattern of uniformity is identified. If the uniformity index under a certain path shows an upward trend over time, it indicates that the photoperiod promotes the enhanced growth stability; if the curve fluctuates violently or continues to decline, it reflects that the photoperiod inhibits stability. At the same time, the slope and fluctuation range of the uniformity index under each path are calculated to quantify the trend intensity and stability differences. The sliding window analysis results are visualized to generate a curve graph of the uniformity index changing over time under each photoperiod path. The horizontal axis is time and the vertical axis is the uniformity index value. Different paths are distinguished by different colors. By comparing the curves, the trend of the influence of the photoperiod on growth stability is presented. The static analysis results of the stem diameter variation coefficient are further combined to verify the reliability of the dynamic trend: if the value of the stem diameter variation coefficient of a certain path is low and the uniformity curve shows an upward trend, then the photoperiod is confirmed to be a high-adaptability path. If the value of the stem diameter variation coefficient is high and the curve fluctuates greatly, it is marked as a low-adaptability path.

[0086] The expression of the uniformity index is as follows:

[0087]

[0088] Where H is the uniformity index, the larger the value, the more uniform the population growth, n is the number of samples, the number of seedlings under the same photoperiod path, i is the sample number, which is used to traverse all seedling individuals, p i is the relative proportion of the stem thickness of the i-th seedling, reflecting the weight of the stem thickness of a single seedling in the total population, x i is the measured stem diameter of the i-th seedling, which is normalized to the interval [0,1]. ln(·) is the natural logarithm (with base e), which is used to quantify the degree of dispersion of the distribution (the essence of entropy).

[0089] The specific work content is: through the Internet of Things sensor network deployed in the target area, the stem thickness, chlorophyll content and corresponding photoperiod parameters are synchronously collected at a fixed interval of every 4 hours to form a multi-dimensional original data stream. After the data is uploaded to the edge computing node, outlier detection and elimination are performed. The statistical method based on the 3σ criterion is used to identify and filter outliers caused by sensor failure or environmental interference. The time series linear interpolation method is used for missing values, and the data mean of adjacent time points is used to fill the gaps to ensure data continuity. The stem thickness and chlorophyll content are normalized to eliminate dimensional differences. The photoperiod parameters are retained as actual values, and finally a standardized time series data set is generated, which includes five-dimensional fields: timestamp, seedling category, stem thickness, chlorophyll content and photoperiod; based on the standardized time series data set, the seedling category is divided according to the clustering results, and further grouped according to the photoperiod path. For each type of seedling stem diameter data under the same photoperiod path, the coefficient of variation (CV) of stem diameter was calculated to quantify the degree of growth dispersion. At the same time, the evenness index (calculated based on the entropy value of stem diameter data) was introduced to evaluate growth consistency. The evenness index was dynamically analyzed using a sliding window (the window width was set to 7 time points), and a time series curve graph was generated to analyze the trend of the impact of photoperiod on growth stability. Combined with the results of the dynamic analysis of evenness, photoperiod paths with a coefficient of variation (CV) of stem diameter less than 15% were selected as candidate solutions, indicating that the seedling growth dispersion was low and the stability was high under this photoperiod. The evenness index of the candidate paths was further tested for time trend to eliminate short-term fluctuation interference and confirm the long-term stability advantage. Finally, the photoperiod paths that met the conditions were marked as highly adaptable paths, and a path-seedling category mapping table was generated to clarify the optimal photoperiod range corresponding to each type of seedling.

[0090] In step 6, the optimization and control process of lighting parameters is as follows:

[0091] Every 24 hours, based on the photoperiod path evaluation results, the stem elongation rate and chlorophyll content of the seedlings under each photoperiod path were calculated. If the path caused the stem elongation rate to be greater than 0.5 cm / d (excessive growth) or the chlorophyll content to be less than 30 SPAD (premature aging), it was marked as a poor-quality path and its pheromone concentration was reduced. Otherwise, it was marked as a high-quality path and the pheromone concentration was maintained or enhanced. The initial pheromone concentration was evenly distributed according to the photoperiod to ensure full coverage. According to the daily evaluation results, the pheromone of the poor-quality path was reduced at a fixed decay rate, and the pheromone of the high-quality path was increased in proportion to the fitness. If a path If a path is marked as poor quality for three consecutive days, it will be forced to be eliminated. If the pheromone ratio of a high-quality path exceeds the threshold of 80%, it will be retained first and its coverage will be expanded. After 10 iterations, only the paths with the top 20% pheromone concentrations will be retained to form a candidate photoperiod set. The candidate photoperiod set will be grouped by seedling category, and the photoperiod distribution of the paths in each group will be counted. The photoperiod range with the highest pheromone concentration will be locked as the optimal matching solution. The optimal matching solution will be input into the lighting control system, and the photoperiod will be dynamically adjusted to ensure that the deviation between the actual parameters and the optimal matching solution is less than 5%, achieving precise control.

[0092] The specific work content is: based on the 24-hour data collected by the Internet of Things sensor network, the stem elongation rate and chlorophyll content of the seedlings are calculated according to the photoperiod path group, and the path quality is determined by the dynamic threshold: if the stem elongation rate exceeds 0.5cm / d (indicating excessive growth) or the chlorophyll content is lower than 30SPAD (indicating premature aging), it is marked as a poor-quality path, otherwise it is marked as a high-quality path. In the initial stage, the pheromone concentration of all photoperiod paths is initialized according to the principle of uniform distribution to ensure coverage of the full range of photoperiods and avoid local optimal solutions due to initial deviations; according to the daily evaluation results, the pheromone concentration of the poor-quality path is reduced at a fixed decay rate (20%), and the pheromone concentration of the high-quality path is increased according to the fitness ratio, where fitness is defined as the product of the inverse of the stem elongation rate and the chlorophyll content, forming a dynamic adjustment mechanism of survival of the fittest. If a path is marked as poor-quality for three consecutive days, it will be forced to be eliminated to avoid Invalid resource occupation: if the pheromone concentration of high-quality paths exceeds the threshold of 80%, its photoperiod coverage will be retained and expanded first to improve the competitiveness of high-quality paths. After 10 iterative updates, only the paths with the top 20% pheromone concentrations will be retained to form a candidate photoperiod set. The positive feedback of pheromone concentration will strengthen the high-quality paths, and the negative feedback will inhibit the low-quality paths, gradually approaching the global optimal solution. The candidate photoperiod set will be grouped according to seedling categories (weak seedlings, medium seedlings, and strong seedlings), and the photoperiod distribution of each group of paths will be counted. The continuous photoperiod range with the highest pheromone concentration will be locked as the optimal matching solution. The optimal matching solution will be input into the lighting control system. By dynamically adjusting the switching time of the LED lights or the opening and closing angles of the sunshade nets, the deviation between the actual photoperiod and the optimal solution will be ensured to be less than 5%, thereby achieving precise environmental control, balancing the growth rate and health status of seedlings, avoiding excessive growth or premature aging, and improving crop yield and quality consistency.

[0093] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A method for dynamically optimizing and controlling environmental parameters for potato seedling grading, characterized in that: The following steps are involved: Step 1: Using IoT sensors deployed in the target area, data on stem diameter, chlorophyll content, and plant height of potato seedlings are collected. K-means clustering is used to classify the seedlings into three categories: weak seedlings, medium seedlings, and strong seedlings. The duration of the initial photoperiod is then set. Step 2: Using stem elongation rate and chlorophyll content as constraints, a gradient photoperiod experiment was conducted on each type of potato seedlings to establish a photoperiod-growth rate response surface model and preliminarily determine the appropriate photoperiod interval; Step 3: Based on the photoperiod-growth rate response surface model, the ant colony algorithm parameters and pheromone matrix are initialized, and the light intensity is introduced through the heuristic factor to optimize the path selection weight; Step 4: The ant colony algorithm selects the photoperiod path by combining pheromone concentration and heuristic factor probability to screen the photoperiod combination; Step 5: Using IoT sensors to collect data on stem diameter and chlorophyll content at fixed intervals, and analyzing the uniformity of seedling growth under each photoperiod path; Step 6: Update pheromones every 24 hours based on the photoperiod path evaluation results, eliminate paths that lead to excessive growth or premature aging, and lock the optimal photoperiod-graded seedling matching scheme after 10 iterations to optimize and control the lighting parameters.

2. The method for dynamically optimizing and controlling environmental parameters for potato seedling grading according to claim 1, wherein: In step 1, the process of setting the duration of the initial light cycle is as follows: IoT sensors deployed in target areas simultaneously collected physiological data on potato seedlings, including stem diameter, chlorophyll content, and plant height. After data cleaning and standardization, a K-means clustering algorithm was used to classify the seedlings into three categories: weak, medium, and strong, based on the Euclidean distance between these physiological indicators. According to the seedling grading results and the comprehensive growth requirements of potato seedlings, the initial photoperiod duration ranges were set for the three types of seedlings. Among them, the initial photoperiod duration ranges for weak seedlings, medium seedlings, and strong seedlings were 10-12h, 12-14h, and 14-16h, respectively.

3. The method for dynamically optimizing and controlling environmental parameters for potato seedling grading according to claim 2, wherein: In step 2, the process of preliminarily determining the appropriate photoperiod interval is as follows: Using stem elongation rate and chlorophyll content as constraints, potato seedlings were grouped according to the results of the previous grading process. A photoperiod gradient of 10-16 hours was set, with a step length of 2 hours. Each group was repeated three times and cultured continuously for 14 days in a controlled environment. The stem elongation rate and chlorophyll content were simultaneously monitored. Based on the experimental data, photoperiod combinations that meet the constraints were screened. With photoperiod and seedling type as independent variables, and stem elongation rate and chlorophyll content as dependent variables, a response surface model was fitted using multiple linear regression to construct a photoperiod-growth rate response surface model to quantify the nonlinear effect of photoperiod on growth indicators. The accuracy of the model was evaluated through cross-validation, the prediction accuracy of the model was verified, the nonlinear relationship between photoperiod and growth rate was output, and the appropriate photoperiod range was preliminarily determined.

4. The method for dynamically optimizing and controlling environmental parameters for potato seedling grading according to claim 3, wherein: The process of constructing the photoperiod-growth rate response surface model and preliminarily determining the appropriate photoperiod interval is as follows: Experimental data on photoperiod and seedling type, as well as the corresponding stem elongation rate and chlorophyll content, were acquired simultaneously. Outliers were removed, continuous variables were Z-score standardized, and categorical variables were kept in dummy coding form and directly included. The acquired data were integrated to obtain a comprehensive dataset, which was divided into training and test sets in a ratio of 7:

3. With photoperiod and seedling type as independent variables, and stem elongation rate and chlorophyll content as dependent variables, a photoperiod-growth rate response surface model was constructed using multiple linear regression. The least squares method was used to fit the model, calculate the regression coefficients and their significance, and screen for significant variables. Three-dimensional response surfaces of photoperiod-growth rate and photoperiod-chlorophyll content were generated based on the regression coefficients, and the marginal effects of photoperiod on growth indicators including stem elongation rate and chlorophyll content were analyzed through partial derivatives. Segmented regression was introduced to deal with nonlinear relationships, and the photoperiod was divided into two segments, <12h and ≥12h. Models were fitted separately, and the AIC values ​​of the segmented model and the global model were compared. When ΔAIC>2, the segmented model was selected and determined as the optimal model. Then, the interval corresponding to the peak value ±0.5 standard deviation of the response surface was taken as the appropriate photoperiod interval for each seedling type.

5. The method for dynamically optimizing and controlling environmental parameters for potato seedling grading according to claim 2, wherein: In step 3, the process of optimizing the path selection weight is as follows: Based on the photoperiod-growth rate response surface model, the initial ant colony algorithm parameters were set, including the number of ants, pheromone volatility coefficient, and initial pheromone concentration. A three-dimensional structure was constructed based on the combination of seedling type and photoperiod to form a pheromone matrix, whose dimensions corresponded to the grid points of the photoperiod-growth rate response surface. Light intensity is introduced into the ant colony algorithm as a heuristic factor, mapped to the interval [0,1] through normalization, and together with pheromone concentration, it constitutes the path selection probability formula, namely the transition probability. The heuristic factor weight is dynamically adjusted so that the guiding effect of light intensity on path selection decays with the increase of iteration number.

6. The method for dynamically optimizing and controlling environmental parameters for potato seedling grading according to claim 5, characterized in that: In step 4, the process of screening the photoperiod combination is as follows: Each ant locates the corresponding level of the pheromone matrix according to the current seedling category, calculates the transfer probability of all feasible neighboring grid points at the current location by combining the pheromone concentration and normalized light intensity, and randomly selects the next location based on the transfer probability; After the ants complete their path selection, they record the photoperiod-seedling combinations they have passed through, and after completing one iteration, they release pheromones along the path they have passed through. Continuously iterate, update the pheromone matrix and evaluate the path adaptability in each round, set a double termination condition, and after the iteration is terminated, extract the grid point with the highest pheromone concentration in the corresponding photoperiod interval for each type of seedling from the final pheromone matrix as its optimal photoperiod matching solution.

7. The method for dynamically optimizing and controlling environmental parameters for potato seedling grading according to claim 6, wherein: The process of determining the optimal photoperiod matching scheme is as follows: The ant colony algorithm continuously iterates and updates the three-dimensional pheromone matrix. In each iteration, after the ant colony completes the path selection based on the current pheromone distribution and light heuristic information, it performs a global update on the three-dimensional pheromone matrix. The pheromone concentration of all grid points is attenuated according to the preset volatility coefficient to simulate the natural volatility process. The pheromone increment is superimposed on the corresponding grid points based on the growth rate prediction value of the path passed by the ants. At the same time, the path adaptability is re-evaluated after each iteration, and the comprehensive adaptability score is calculated. The growth rate prediction value of the photoperiod-seedling combination is associated with the pheromone concentration, gradually guiding the search direction to converge to the high-fitness area to determine the optimal path. The algorithm sets dual termination conditions, namely, a preset maximum number of iterations and dynamic convergence judgment. The preset maximum number of iterations prevents infinite loops; through dynamic convergence judgment, when the concentration value of the optimal path does not change significantly after 10 consecutive iterations, the algorithm is considered to have converged to a stable solution. After the algorithm terminates, the optimal photoperiod matching scheme for each type of seedling is extracted from the final pheromone matrix. Weak seedlings are locked in the long-day interval, strong seedlings are determined to be in the short-day interval, and medium seedlings are matched with the intermediate photoperiod.

8. The method for dynamically optimizing and controlling environmental parameters for potato seedling grading according to claim 6, wherein: In step 5, the process of analyzing the uniformity of seedling growth under each photoperiod path is as follows: IoT sensors in the target area collect data on stem diameter and chlorophyll content at a fixed interval of every 4 hours, and simultaneously record the corresponding photoperiod parameters. After the data is uploaded to the edge computing node, outliers are removed and missing values ​​are interpolated to generate a standardized time series dataset. Based on the clustered seedling categories, the seedlings were grouped by photoperiod path. The coefficient of variation of stem diameter and evenness index of each seedling type under the same photoperiod path were calculated. The temporal dynamics of evenness were analyzed using a sliding window, and a curve chart of evenness changes over time under each path was generated. Combined with the evenness analysis results, photoperiod paths with a stem diameter variation coefficient of <15% were screened as candidate options and marked as high-adaptability paths.

9. The method for dynamically optimizing and controlling environmental parameters for potato seedling grading according to claim 8, characterized in that: The process of analyzing the time dynamics of uniformity through a sliding window and generating a curve graph of uniformity changes over time for each path is as follows: Based on the seedling classification results obtained by K-means clustering, the data were grouped according to the photoperiod parameters. For each group of data, the coefficient of variation of stem diameter was calculated, that is, the ratio of the standard deviation of stem diameter to the mean, to quantify the degree of dispersion of stem diameter among seedlings under the same photoperiod. At the same time, the evenness index was introduced to evaluate growth stability. The evenness index was calculated based on the entropy value of the stem diameter data. The sliding window method was used to analyze the temporal trend of the evenness index. The window width was set to a fixed time interval, and the mean of the evenness index within each window was calculated to generate time series data. By comparing the time series curves under different photoperiod paths, the variation pattern of evenness was identified. At the same time, the slope and fluctuation range of the evenness index under each path were calculated to quantify the difference in trend strength and stability. The sliding window analysis results were visualized to generate a graph showing the change of the evenness index over time under each photoperiod path. The horizontal axis was time and the vertical axis was the evenness index value. Different paths were distinguished by different colors. By comparing the curves, the trend of the impact of the photoperiod on growth stability was presented.

10. The method for dynamically optimizing and controlling environmental parameters for potato seedling grading according to claim 8, characterized in that: In step 6, the optimization and control process of the lighting parameters is as follows: Based on the photoperiod path evaluation results, the stem elongation rate and chlorophyll content of the seedlings under each photoperiod path were calculated every 24 hours. If the path resulted in a stem elongation rate greater than 0.5 cm / d or a chlorophyll content less than 30 SPAD, it was marked as a poor-quality path, otherwise it was marked as a high-quality path. According to the daily evaluation results, the pheromone of the poor-quality paths is reduced at a fixed decay rate, and the pheromone of the high-quality paths is increased according to the fitness ratio. If a path is marked as poor-quality for three consecutive days, it is forced to be eliminated. If the pheromone ratio of the high-quality path exceeds the threshold of 80%, it is retained and its coverage is expanded. After 10 iterations, only the paths with the top 20% pheromone concentration are retained to form the candidate photoperiod set; The candidate photoperiod set is grouped by seedling category, the photoperiod distribution of each group of paths is counted, the photoperiod range with the highest pheromone concentration is locked as the optimal matching solution, the optimal matching solution is input into the lighting control system, and the photoperiod is dynamically adjusted.

Images