A Modeling and Growth Prediction Method for Tomato Seedling Models

Abstract

The present invention relates to a method for modeling and growth prediction of tomato seedling models. This method comprehensively considers the effects of temperature, light, and light-temperature coupling on the development of tomato seedlings, and establishes a growth and development prediction model for tomato morphological indicators and strong seedling index based on photosynthetically active temperature sum and relative light effect. By processing and calculating photosynthetically active radiation, this method eliminates the dimension of light data, defines a new parameter of relative light effect to describe the light energy absorbed by crops, reduces the magnitude difference between effective temperature sum and photosynthetically active radiation, and at the same time considers the night growth and development process in the model, thereby optimizing the model prediction accuracy. The trend of the tomato seedling model established by this method conforms to the seedling growth law, the model fitting degree is high, and it can accurately predict the change process of tomato seedling morphological indicators and strong seedling index.

Description

Technical Field

[0001] The present invention belongs to the fields of agricultural informatization and automation, and particularly relates to a method for modeling a tomato seedling model and predicting growth. Background Art

[0002] The research on greenhouse crop growth and development models started in the late 1970s and early 1980s with the large-scale cultivation of greenhouse crops. With the improvement of greenhouse management technology, the development of Internet of Things sensor technology, and the continuous in-depth research on the growth and development mechanism of horticultural crops, there are more and more studies on greenhouse crop growth models currently. Today, with the rapid development of agricultural information technology, crop growth models are considered the scientific basis and core technology of "digital agriculture" and the bridge to agricultural informatization, and have become a popular research field at home and abroad.

[0003] During the factory seedling raising process, if the environmental conditions are not properly controlled, the seedlings are extremely prone to leggy growth, which in turn affects the growth, development, yield and quality of tomatoes. Temperature and light in the greenhouse are important environmental factors affecting seedling growth. By studying crop growth models and then realizing the precise control of environmental conditions such as greenhouse temperature and light, it is of great significance for cultivating strong seedlings.

[0004] Although many growth models of horticultural crops have been studied in China at present, most of these models are plant whole growth period models. Among them, for seedling production, there are few studies on growth models focusing on the seedling stage of crops, and even fewer on growth models related to the tomato seedling stage. The morphological indexes and strong seedling index of seedlings are the most intuitive parameters for evaluating the health of seedlings, but there are few reports on the simulation model of tomato strong seedling index. The existing growth models mainly include morphological index models, dry matter models, growth and development days models, etc. Although some growth models have certain reference value for the growth trend of crops, the environmental factors included for reference are relatively single.

[0005] In order to establish a more accurate growth model, researchers have begun to focus on models involving multiple environmental factors. Although some strong seedling index models comprehensively consider the influence of multiple environmental factors on crop growth and use the product of each environmental factor as the independent variable for modeling, since the value of photosynthetically active radiation at night is zero, such models ignore the growth of seedlings at night. Research shows that although the day temperature and daytime light intensity are kept the same, only by changing the night growth temperature of seedlings, there are significant differences in the growth and development of seedlings under different night temperature conditions. This further shows that there is also growth and development of seedlings at night when the photosynthetically active radiation is 0, while the existing radiation heat sum model ignores the growth of seedlings at night. Summary of the Invention

[0006] (1) Technical Problems to be Solved

[0007] In order to overcome the deficiencies of the above-mentioned prior art, specifically, to eliminate the single environmental factor involved in the growth model, and the problem that the model established based on radiant heat product ignores the nighttime growth of the plant, to avoid model errors caused by too few environmental parameters and omission of the nighttime growth and development process of the plant, and to improve the accuracy and prediction ability of the model, the present invention provides a tomato seedling model building and growth prediction method.

[0008] (II) Technical solution

[0009] In order to solve the above problems, the present invention proposes a tomato seedling model building and growth prediction method, and the technical solution is as follows.

[0010] A tomato seedling modeling and growth prediction method, characterized in that a tomato seedling growth prediction model is established based on effective accumulated temperature and relative light effect to predict the growth and development of the seedlings, specifically comprising the following steps:

[0011] S1, obtain the environmental parameters in the greenhouse during the seedling raising process, including temperature and photosynthetically active radiation;

[0012] S2, covering the physiological age of seedlings, sampling was performed 8 times to obtain the phenotypic parameters of tomato plug seedlings, including plant height, stem diameter, leaf area, dry weight and seedling index;

[0013] S3, process the environmental parameters, eliminate the dimension and consider the plant growth and development characteristics, and calculate the effective accumulated temperature and relative light effect;

[0014] S4, fit the calculated effective accumulated temperature and relative light effect with the obtained seedling phenotypic parameters, and compare R 2 and RMSE, determine the model equation;

[0015] S5, using the light and temperature data of the weather forecast for the next few days as input, accurately predicting the growth and development of the tomato seedlings based on the model equation determined in step S4.

[0016] Preferably, in step S1, environmental parameters in the seedling raising process are obtained, specifically, environmental parameters of the entire growth period of the seedlings are automatically collected by sensors; the collected data include the air temperature above the seedling canopy and the total photosynthetic radiation; the collection frequency is once every 10 minutes.

[0017] Preferably, in step S2, the phenotypic parameters of tomato plug seedlings are obtained. Specifically, the tomato seedling raising period is sampled 8 times according to the physiological seedling age of the seedling growth and development, which are the cotyledon flattening stage, the 1-leaf 1-heart stage, the 2-leaf 1-heart stage, 7 days after the 2-leaf 1-heart stage, the 3-leaf 1-heart stage, the 4-leaf 1-heart stage, 2 days after the 4-leaf 1-heart stage, and the 5-leaf 1-heart stage; 30 seedlings are randomly selected for each sampling to measure the morphological indexes; the measurement indexes include plant height, stem diameter, leaf area, total plant dry weight, and strong seedling index; the strong seedling index is calculated according to formula (1):

[0018] Strong seedling index = stem diameter / plant height × total plant dry weight (1).

[0019] Preferably, when processing the environmental parameters in step S3, the effective accumulated temperature is calculated by using formulas (2-1), (2-2), and (2-3).

[0020] GD day =T mean -T b (2-1)

[0021] GDD day =GDD day-1 +GD day (2-2)

[0022]

[0023] In the formula, GD day is the effective temperature for a certain purpose, with the unit of ℃·d; GDD day-1 is the effective temperature of the previous day, with the unit of ℃·d; GDD day is the effective accumulated temperature from the previous growth period to the current, with the unit of ℃·d, T mean is the actually measured daily average temperature, T b 、T m are the lower biological limit temperature and the upper biological limit temperature for the development of tomato at a certain development stage respectively.

[0024] Preferably, when calculating the effective accumulated temperature, different settings are adopted for the lower limit temperature and the upper limit temperature of the tomato seedlings during the germination period and the seedling stage; specifically, the lower growth limit temperature during the germination period is set at 12℃, the upper growth limit temperature is set at 35℃, the lower growth limit temperature during the seedling stage is set at 10℃, and the upper growth limit temperature is set at 35℃.

[0025] Preferably, when processing the environmental parameters in step S3, the relative light effect is calculated according to formulas (3-1) and (3-2).

[0026] RLE=∑DRLE (3-1)

[0027]

[0028] In the formula, DRLE (k) represents the relative light effect on the kth day, and L min represents the light compensation point of tomato seedlings, with the unit of μmol·m -2 ·s -1 , that is, at this time, the photosynthetic rate and respiration rate of tomato seedlings are equal, and no dry matter is accumulated. L max represents the light saturation point of tomato seedlings, with the unit of μmol·m -2 ·s -1 , that is, the photosynthetic rate of tomato seedlings will no longer increase with the increase of light intensity. L is the actually measured average photosynthetically active radiation per hour, with the unit of μmol·m -2 ·s -1 .

[0029] Preferably, when calculating the relative light effect, the light compensation point of tomato seedlings is set to 37.05 μmol·m -2 ·s -1 , and the light saturation point of tomato seedlings is set to 1361.49 μmol·m -2 ·s -1 .

[0030] Preferably, in step S4, the environmental parameters and seedling phenotype parameters are fitted to determine the model equation. Specifically, the lsqcurvefit function of Matlab is used to perform binary nonlinear curve fitting on the data, and the model parameters are determined by comparing R 2 and RMSE. Formulas (4-1) and (4-2) are as follows:

[0031]

[0032] In formula (4-1), SP is the sum of products; SS x is the mean sum of squares of deviations of x from the mean; SS y is the mean sum of squares of deviations of y from the mean.

[0033]

[0034] In formula (4-2), OBSi is the observed value, SIMi is the simulated value, i is the sample number, and n is the sample size.

[0035] Preferably, step S5 accurately predicts the growth and development of tomato seedlings. Specifically, through the weather forecast for the next few days, the effective accumulated temperature and relative light effect for the next few days are calculated, and according to the model equation determined in S4, the phenotype parameters of tomato seedlings for the next few days are calculated to accurately predict the time when the seedlings are ready for outplanting.

[0036] (III) Beneficial effects

[0037] Compared with the prior art, the tomato seedling model modeling and growth prediction method provided by the present invention has significantly positive technical effects, which are specifically manifested in the following aspects:

[0038] (1) The present invention not only considers the effects of temperature and light on seedlings respectively, but also considers the effect of light-temperature coupling, and sets multiple parameters to fit the binary curve model, which is closer to the actual growth of seedlings in terms of physical meaning representation, significantly superior to the prior art, and helps to obtain a prediction model with higher accuracy.

[0039] (2) The present invention eliminates the dimension of light data through mathematical calculation, and uniquely defines the parameter of relative light effect to describe the light energy absorbed by crops, effectively reducing the magnitude difference between effective accumulated temperature and photosynthetically active radiation, and further optimizing the accuracy of model fitting and prediction.

[0040] (3) The present invention fits the phenotypes and strong seedling index of tomato seedlings. The results show that the R of the fitting equations of each phenotype 2 is greater than 0.99, indicating that its fitting effect is excellent, significantly superior to the prior art. The three-dimensional simulation composite maps of each phenotype established with this model style conform to the growth law of seedlings. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is the technical flow chart of the present invention;

[0042] Figure 2 is the cultivation environment temperature diagram of Examples 1-2;

[0043] Figure 3 is the cultivation environment light diagram of Examples 1-2;

[0044] Figure 4 is the fitting model diagram of plant height, stem diameter, and leaf area of Example 1;

[0045] Figure 5 is the fitting model diagram of aboveground dry weight, underground dry weight, and strong seedling index of Example 1;

[0046] Figure 6 is the relationship diagram between simulated values and observed values of each phenotype of Example 2. DETAILED DESCRIPTION OF THE INVENTION

[0047] The present invention will be further described below with reference to the drawings and implementation cases. The examples cited are only used to explain the present invention and are not intended to limit the scope of the present invention.

[0048] Both Example 1 and Example 2 of the present invention adopt the Figure 1 technical flow chart as shown.

[0049] In Example 1 and Example 2, the tomato seeds used were Qidali (purchased from Syngenta Seeds Co., Ltd.) and Jinpeng No. 8 (purchased from Xi'an Jinpeng Seedlings Co., Ltd.), respectively. The substrate was composed of peat:vermiculite:perlite with a volume ratio of 3:1:1. The tomato seedlings were cultivated in plug trays, which were XQA72 plug trays (purchased from Taizhou Plastics Co., Ltd.), and the volume of each hole was 36 ml. During the cultivation of tomato seedlings, two fertilizers were alternately used for topdressing, namely Larui Special Fertilizer No. 1 and No. 2 (purchased from Greencare Company). The sensor used for collecting environmental data in the greenhouse was the ZENTRA sensor.

[0050] In Example 1, the tomato seeds used were Qidali, and in Example 2, the tomato seeds used were Jinpeng No. 8. The cultivation environmental temperatures in Example 1 and Example 2 were as Figure 2 shown, and the cultivation environmental light was as Figure 3 shown. The prediction of the growth and development of the seedlings in both examples included the following steps:

[0051] S1. Obtain the environmental parameters in the greenhouse during the seedling cultivation process, including temperature and photosynthetically active radiation;

[0052] S2. Cover the physiological seedling age of the seedlings, take samples 8 times, and obtain the phenotypic parameters of the tomato plug seedlings, specifically including plant height, stem diameter, leaf area, dry weight, and strong seedling index;

[0053] S3. Process the environmental parameters, eliminate the dimension and consider the characteristics of plant growth and development, and calculate the effective accumulated temperature and relative light effect;

[0054] S4. Fit the calculated effective accumulated temperature and relative light effect with the obtained seedling phenotypic parameters, compare R 2 and RMSE, and determine the model equation;

[0055] S5. Use the light and temperature data of the weather forecast for the next few days as the input, and accurately predict the growth and development of tomato seedlings based on the model equation determined in step S4.

[0056] In step S1, the environmental parameters during the seedling cultivation process were obtained specifically by automatically collecting the environmental parameters of the whole growth period of the seedlings through sensors; the collected data included the air temperature and total photosynthetic radiation above the seedling canopy; the collection frequency was to collect data once every 10 minutes.

[0057] In step S2, the phenotypic parameters of tomato plug seedlings are obtained. Specifically, the tomato seedling raising period is sampled 8 times according to the physiological seedling age of seedling growth and development, which are the cotyledon flattening stage, the 1-leaf 1-heart stage, the 2-leaf 1-heart stage, 7 days after the 2-leaf 1-heart stage, the 3-leaf 1-heart stage, the 4-leaf 1-heart stage, 2 days after the 4-leaf 1-heart stage, and the 5-leaf 1-heart stage; 30 seedlings are randomly selected for each sampling to measure the morphological indexes; the measurement indexes include plant height, stem diameter, leaf area, total plant dry weight, and strong seedling index; the strong seedling index is calculated according to formula (1):

[0058] Strong seedling index = stem diameter / plant height × total plant dry weight (1).

[0059] When processing the environmental parameters in step S3, the effective accumulated temperature is calculated using formulas (2-1), (2-2), and (2-3).

[0060] GD day = T mean - T b (2-1)

[0061] GDD day = GDD day - 1 + GD day (2-2)

[0062]

[0063] In the formula, GD day is the effective temperature of a certain day, with the unit of ℃·d; GDD day-1 is the effective temperature of the previous day, with the unit of ℃·d; GDD day is the effective accumulated temperature from the previous growth period to the current, with the unit of ℃·d, and T mean is the actually measured daily average temperature, and T b , T m are the biological lower limit temperature and upper limit temperature for the development of tomatoes at a certain development stage, respectively.

[0064] When calculating the effective accumulated temperature, different settings are used for the lower limit temperature and upper limit temperature of tomato seedling development during the germination period and the seedling stage; specifically, the lower limit temperature for growth during the germination period is set at 12℃, and the upper limit temperature for growth is set at 35℃. The lower limit temperature for growth during the seedling stage is set at 10℃, and the upper limit temperature for growth is set at 35℃.

[0065] When processing the environmental parameters in step S3, the relative light effect is calculated according to formulas (3-1) and (3-2).

[0066] RLE = ∑DRLE (3-1)

[0067]

[0068] In the formula, DRLE (k) represents the relative light effect on the k-th day, and L min represents the light compensation point of tomato seedlings, with the unit of μmol·m -2 ·s -1 , that is, at this time, the photosynthetic rate and respiration rate of tomato seedlings are equal, and no dry matter is accumulated. L max represents the light saturation point of tomato seedlings, with the unit of μmol·m -2 ·s -1 , that is, the photosynthetic rate of tomato seedlings will no longer increase with the increase of light intensity. L is the actually measured average photosynthetically active radiation per hour, with the unit of μmol·m -2 ·s -1 .

[0069] When calculating the relative light effect, the light compensation point of tomato seedlings is set to 37.05 μmol·m -2 ·s -1 , and the light saturation point of tomato seedlings is set to 1361.49 μmol·m -2 ·s -1 .

[0070] In step S4, the environmental parameters and seedling phenotype parameters are fitted to determine the model equation. Specifically, the lsqcurvefit function of Matlab is used to perform binary nonlinear curve fitting on the data, and the model parameters are determined by comparing R2 and RMSE. Formulas (4-1) and (4-2) are as follows:

[0071]

[0072] In formula (4-1), SP is the sum of products; SS x is the mean sum of squares of deviations of x from the mean; SS y is the mean sum of squares of deviations of y from the mean.

[0073]

[0074] In formula (4-2), OBSi is the observed value, SIMi is the simulated value, i is the sample number, and n is the sample size.

[0075] Step S5 accurately predicts the growth and development of tomato seedlings. Specifically, through the weather forecast for the next few days, the effective accumulated temperature and relative light effect for the next few days are calculated, and according to the model equation determined in S4, the phenotype parameters of tomato seedlings for the next few days are calculated to accurately predict the time when the seedlings are ready to be transplanted.

[0076] The statistical results of the phenotype indicators and environmental parameters in Example 1 are shown in Table 1.

[0077] Table 1 Statistical results of phenotype indicators and environmental parameters in Example 1

[0078]

[0079] It should be added that the numerical values of each index in the table are the averages of 30 samples; '--' indicates that the data cannot be measured because the seedlings are too small, and the same applies hereinafter.

[0080] Example 1 uses a model style in the form of to fit the phenotypes and strong seedling index of tomato seedlings. The fitting results of plant height, stem diameter, and leaf area are shown in Figure 4 , and the fitting results of aboveground dry weight, underground dry weight, and strong seedling index are shown in Figure 5 . The fitting model is shown in Table 2. The coefficient of determination R 2 is preferably close to 1; the smaller the RMSE value, the better the consistency between the simulated value and the measured value, and the smaller the deviation between the simulated value and the measured value, that is, the more accurate and reliable the simulation result of the model. Generally, it is considered excellent within 15. Calculate R 2 and RMSE according to formulas (4-1) and (4-2). The calculation results are shown in Table 2. The R 2 of each phenotype fitting model is greater than 0.99, and the RMSE value is very small, indicating that the model fitting effect is good.

[0081] Table 2 Parameters and effects of the phenotype fitting model in Example 1

[0082]

[0083]

[0084] Using the established model (Table 2), the simulated values are calculated according to the environmental data collected during the seedling raising process in Example 2 ( Figure 2 ), and the coefficient of determination R 2 and root mean square error RMSE are used to statistically analyze the degree of agreement between the model value SIM and the observed value OBS. The simulated values and observed values of Example 2 calculated according to the model are shown in Table 3, and the environmental parameters are shown in Table 4.

[0085] Table 3 Simulated values and observed values of indicators in Example 2

[0086]

[0087] Table 4 Statistical results of environmental parameters in Example 2

[0088]

[0089] Use the RMSE model to test the statistical method (Equation (4-2)) for statistical analysis. The RMSE values of the plant height, stem diameter, leaf area, aboveground dry weight, underground dry weight, and strong seedling index models are calculated to be 0.2776, 0.0452, 2.3947, 0.0089, 0.0012, and 0.0024, respectively. The calculation results show that the predictive performance of each phenotypic model is good.

[0090] Make a 1:1 relationship graph of the simulated values and observed values in Example 2 and perform regression analysis. The regression analysis results of each phenotypic model are shown in Table 4 and Figure 6 as follows. By analyzing the equations of each phenotypic simulated value and observed value, the results show that the R 2 of each equation is greater than 0.9, indicating that the relationship between each simulated value and observed value is close.

[0091] Table 4 Tomato Seedling Phenotypic Fitting Model

[0092]

[0093] The specific examples described in the application are only illustrative of the spirit of the present invention. Those skilled in the art to which the present invention pertains can make various modifications or supplements to the specific examples described in the present invention, or use alternative methods of the same type, but will not deviate from the spirit of the present invention or exceed the scope defined by the appended claims.

Claims

1. A method for modeling a tomato seedling model and predicting growth, characterized in that, A tomato seedling growth model is established based on effective accumulated temperature and relative light effect to predict the growth and development of seedlings. The specific steps are as follows: S1. Obtain the environmental parameters in the greenhouse during the seedling raising process, including temperature and photosynthetically active radiation; S2. Cover the physiological seedling age of the seedlings and take samples 8 times to obtain the phenotypic parameters of tomato plug seedlings, specifically including plant height, stem diameter, leaf area, dry weight, and strong seedling index; S3. Process the environmental parameters, eliminate the dimension and consider the characteristics of plant growth and development, and calculate the effective accumulated temperature and relative light effect; S4. Fit the calculated effective accumulated temperature and relative light effect with the seedling phenotype parameters obtained by sampling, and compare R 2 and RMSE to determine the model equation; S5. Use the light and temperature data of the weather forecast for the next few days as input, and accurately predict the growth and development of tomato seedlings based on the model equation determined in step S4; Among them, when processing the environmental parameters in step S3, the effective accumulated temperature is calculated using formulas (2-1), (2-2), and (2-3). (2-1) (2-1) (2-3) Wherein, GD day is the effective temperature of a certain day, with the unit of ℃·d; GDD day-1 is the effective temperature of the previous day, with the unit of ℃·d; GDD day is the effective accumulated temperature from the previous growth stage to the current stage, with the unit of ℃·d, T mean is the actually measured daily average temperature, T b and T m are respectively the biological lower limit temperature and upper limit temperature for the development of tomatoes at a certain development stage; When processing the environmental parameters in step S3, the relative light effect is calculated according to formulas (3-1) and (3-2). （3-1） (3-2) In the formula, DRLE （k） represents the relative light effect on the k-th day, L min represents the light compensation point of tomato seedlings, with the unit of μmol·m -2 ·s -1 , that is, at this time, the photosynthetic rate and respiratory rate of tomato seedlings are equal, and no dry matter is accumulated. L max represents the light saturation point of tomato seedlings, with the unit of μmol·m -2 ·s -1 , that is, the photosynthetic rate of tomato seedlings will no longer increase with the increase of light intensity. L is the actually measured average photosynthetically active radiation per hour, with the unit of μmol·m -2 ·s -1 ; In step S4, the environmental parameters and the seedling phenotypic parameters are fitted to determine the model equation. Specifically, the lsqcurvefit function of Matlab is used to perform binary nonlinear curve fitting on the data, and the model parameters are determined by comparing R 2 and RMSE; Formulas (4-1) and (4-2) are as follows: (4-1) In Equation (4-1), SP is the sum of products; SS x is the mean sum of squares of deviations of x from its mean; SS y is the mean sum of squares of deviations of y from its mean; (4-2) In formula (4-2), OBSi is the observed value, SIMi is the simulated value, i is the sample serial number, and n is the sample size.

2. The method for modeling and growth prediction of tomato seedling models according to claim 1, wherein In step S1, the environmental parameters during the seedling raising process are obtained. Specifically, the environmental parameters of the whole growth period of the seedlings are automatically collected by sensors; the collected data includes the air temperature above the seedling canopy and the total photosynthetic radiation; the collection frequency is to collect data once every 10 minutes.

3. The tomato seedling model modeling and growth prediction method according to claim 1, characterized in that, In step S2, the phenotypic parameters of tomato plug seedlings are obtained. Specifically, the tomato seedling raising cycle is divided into 8 times of sampling according to the physiological seedling age of the seedling growth and development, namely the cotyledon flattening stage, the 1-leaf 1-heart stage, the 2-leaf 1-heart stage, 7 days after the 2-leaf 1-heart stage, the 3-leaf 1-heart stage, the 4-leaf 1-heart stage, 2 days after the 4-leaf 1-heart stage, and the 5-leaf 1-heart stage; 30 seedlings are randomly selected for each sampling to measure the morphological indexes; the measurement indexes include plant height, stem diameter, leaf area, total plant dry weight, and strong seedling index; the strong seedling index is calculated according to formula (1): Strong seedling index = stem diameter / plant height × total plant dry weight (1).

4. The method for modeling and growth prediction of tomato seedling models according to claim 1, characterized in that, When calculating the effective accumulated temperature, different settings are adopted for the lower limit temperature and the upper limit temperature of tomato seedling development in the germination period and the seedling stage; specifically, the lower limit temperature for growth in the germination period is set at 12 °C, the upper limit temperature for growth is set at 35 °C, the lower limit temperature for growth in the seedling stage is set at 10 °C, and the upper limit temperature for growth is set at 35 °C.

5. The method for modeling and growth prediction of tomato seedling models according to claim 1, characterized in that When calculating the relative light effect, the light compensation point of tomato seedlings was set at 37.05 μmol·m -2 ·s -1 , and the light saturation point of tomato seedlings was set at 1361.49 μmol·m -2 ·s -1 .

6. The tomato seedling model modeling and growth prediction method according to claim 1, characterized in that Step S5 accurately predicts the growth and development of tomato seedlings. Specifically, through the weather forecast for the next few days, the effective accumulated temperature and relative light effect for the next few days are calculated, and according to the model equation determined in S4, the phenotypic parameters of tomato seedlings for the next few days are calculated to accurately predict the time for the seedlings to leave the nursery.

Images