Method and device for constructing flowering phase and quality prediction model of cut-flower roses' Jingwan Mountain ', and storage medium

Abstract

The invention relates to the technical field of computers, and particularly discloses a method and a device for constructing a flowering phase and quality prediction model of cut-flower roses' Jingwan Mountain ', and a storage medium, and the method comprises the steps: collecting the internal environment data of a greenhouse for planting the cut-flower roses' Jingwan Mountain 'through a sensor with the Internet of Things; further subdividing and processing the internal environment data collected by the sensor into a plurality of environment factors to obtain an input value; carrying out source-library proportion processing on the cut-flower rosa chinensis' Hongqiwan mountain 'to obtain an output value; each output value and all the input values are independently and randomly sampled multiple times with replacement, and a plurality of data sets containing different input values and corresponding output values are generated; using a random forest algorithm to obtain input values having high correlation with each output value, namely characteristic values; and solving a relation formula of each output value and the related characteristic value through polynomial fitting, so as to obtain the flowering phase and quality prediction model.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method, device and storage medium for constructing a flowering period and quality prediction model for the cut rose "Peach Snow Mountain". Background Art

[0002] Traditional rose cultivation is mainly open-air planting. With the introduction of greenhouses and soilless cultivation models, producers can achieve high-quality, high-yield, high-efficiency, low-energy consumption and environmentally friendly cut flower production under controllable environmental conditions, effective energy use, and reusable and precisely controlled water and fertilizer irrigation. At the same time, this also makes traditional flower farmers face the challenges of theoretical knowledge update and production model transformation. Many flower farmers have rich experience in rose cultivation and management, and most of their experience is "feeling" or "estimated", without specific values ​​or numerical ranges, that is, the master control system and automated equipment under the new production model cannot directly identify and use it. However, the new model has been widely promoted and applied. Cut flower production under the new production model has become an inevitable trend. Traditional flower farmers must accept and gradually adapt and transform. As a part of the new model, environmental control is also a regular setting item in the daily production process, which needs to be adjusted according to market price changes, planting plans, production density, variety types, etc. "Mitao Snow Mountain" is one of the cut rose varieties that is most planted by producers and has the largest market circulation. Therefore, it is necessary to develop a usable computing model for "Mitao Snow Mountain" that can use the real-time changing environmental data of the greenhouse to predict plant growth and development responses, flowering period and quality, and assist in environmental control settings. Summary of the invention

[0003] The purpose of this application is to provide a method, device and storage medium for constructing a flowering period and quality prediction model for the cut rose "Peach Snow Mountain", so as to solve the problem in the prior art of needing to develop a usable computing model for "Peach Snow Mountain" that can utilize real-time greenhouse environmental data to predict plant growth and development responses, flowering period and quality, and assist in environmental control settings.

[0004] To achieve the above-mentioned purpose, the embodiment of the present application provides a method for constructing a flowering period and quality prediction model of the cut rose "Peach Snow Mountain", comprising the following steps: collecting internal environmental data of a greenhouse where the cut rose "Peach Snow Mountain" is planted through a sensor with an Internet of Things;

[0005] Further subdividing and processing the internal environment data collected by the sensor into multiple environmental factors to obtain input values;

[0006] The source-sink ratio of the cut rose "Peach Snow Mountain" was processed to obtain the output value;

[0007] Generate multiple data sets including different input values ​​and corresponding output values ​​by sampling each of the output values ​​and all the input values ​​independently and randomly with replacement multiple times;

[0008] The random forest algorithm is used to obtain input values ​​that have a high correlation with each output value, namely, the eigenvalues;

[0009] The relationship formula between each output value and its related characteristic value is solved by polynomial fitting, so as to obtain the flowering period and quality prediction model.

[0010] Optionally, the internal environment data acquired by the sensor is data information of the top of the vertical flower branch and the surface of the nutrient branch;

[0011] The internal environment data includes at least one of photosynthetically active radiation intensity, temperature, relative humidity and carbon dioxide concentration.

[0012] Optionally, the input values ​​include at least one of the daytime average temperature, the nighttime average temperature, the daytime average humidity, the nighttime average humidity, the photosynthetically active radiation intensity, the carbon dioxide concentration, the relative radiant product and the radiant product.

[0013] Optionally, the source-sink ratio processing includes: obtaining 3 flower branches and 2 leaf area index nutrient branch and foliage curtains, 3 flower branches and 4 leaf area index nutrient branch and foliage curtains, 5 flower branches and 2 leaf area index nutrient branch and foliage curtains, and 5 flower branches and 4 leaf area index nutrient branch and foliage curtains, recording the dates of each growth period during the growth of the flower branches, including the germination period, budding period, and harvesting period, measuring the appearance quality at the time of harvest, and obtaining the output value, which specifically includes: at least one of the flower branch length, flower branch thickness, flower bud height, leaf area and stem, leaf and flower dry weight.

[0014] Optionally, the obtaining the characteristic value by using a random forest algorithm specifically includes:

[0015] The influence of the environmental factor is calculated by summing the reduction in impurity caused by the feature in the node splitting of each decision tree;

[0016] The permutation importance assessment method was used to further verify the impact of features.

[0017] Optionally, solving the relationship formula between each output value and its related characteristic value by polynomial fitting to obtain the flowering period and quality prediction model specifically includes:

[0018] Initialize the polynomial regression prediction model, assuming that there is growth prediction features and merge all the features into one vector, i.e. , and at the same time set the initial parameter vector according to the number of features ,vector The initial values ​​of all parameter values ​​in are random values. The parameter vector is trained and adjusted according to the real data of the prediction target and feature value. The initial polynomial regression prediction model is:

[0019] ,

[0020] In the formula, Represents the predicted value, i.e., the output value, including flower branch thickness, flower branch length, leaf area, bud height, stem weight, and flower dry weight;

[0021] represents the i-th eigenvalue;

[0022] w i It represents the parameter corresponding to the i-th feature associated with the predicted target, and w0 is the intercept term;

[0023] m represents the power of the eigenvalue.

[0024] Optionally, if the degree of the fitted polynomial is high, w can be refitted by ridge regression i Regularization is performed, that is, w i After differentiation, the result approaches 0.

[0025] Optionally, after obtaining the flowering period and quality prediction model, the method further includes:

[0026] Ensure that the cut rose "Peach Snow Mountain" in the greenhouse is planted in the same planting cycle with a spacing of 15 cm, a row spacing of 15 cm, a substrate trough spacing of 80 cm, and a leaf area index of 2 or 4, obtain the internal environment data of the greenhouse at a certain time interval, and further subdivide and process it to obtain characteristic values, and substitute the processed characteristic values ​​into the flowering period and quality prediction model to obtain daily prediction values, that is, output values;

[0027] After obtaining the prediction results of flowering period and quality, it also includes:

[0028] Obtain characteristic values ​​for each predicted value under corresponding source-storage conditions, and perform environmental control settings based on the characteristic values ​​and actual production conditions.

[0029] To achieve the above-mentioned purpose, the present application also provides a device for constructing a flowering period and quality prediction model of a cut rose "Peach Snow Mountain", comprising: a memory; and

[0030] A processor connected to the memory is configured to execute the steps of the method described above.

[0031] To achieve the above objectives, the present application also provides a computer storage medium on which a computer program is stored, wherein the computer program implements the steps of the above method when executed by a machine.

[0032] The embodiments of the present application have the following advantages:

[0033] The flowering period and quality prediction model constructed by the above method can achieve repeated updates of predictions and efficient adjustment of relevant environmental influencing factors. It can also assist in formulating environmental control setting plans according to special market sales targets. In addition, in the process of building the model using the above method, it is possible to supplement the common source-sink ratio scenarios of different varieties in production, so that the model can be used in actual applications, by combining the production density and source-sink quantity of field surveys, to make the prediction closer to the actual situation and more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the implementation methods of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for the implementation methods or the description of the prior art. Obviously, the drawings described below are only exemplary, and for ordinary technicians in this field, other implementation drawings can be derived from the provided drawings without creative work.

[0035] Figure 1 A flowchart of a method for constructing a flowering period and quality prediction model for the cut rose "Peach Snow Mountain" provided in at least one embodiment of the present application;

[0036] Figure 2 A logic block diagram of a model construction method for constructing a cut rose "Peach Snow Mountain" flowering period and quality prediction model provided by at least one embodiment of the present application;

[0037] Figure 3 A module block diagram of a device for constructing a flowering period and quality prediction model for the cut rose "Peach Snow Mountain" provided in at least one embodiment of the present application. DETAILED DESCRIPTION

[0038] The following is a description of the implementation of the present application by specific specific embodiments. People familiar with the technology can easily understand other advantages and effects of the present application from the contents disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0039] It should be noted that the steps in the claims and description of the present application may be executed substantially in parallel or in reverse order under appropriate circumstances, depending on the functions involved.

[0040] In addition, the technical features involved in the different embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0041] At present, there is no model tool for greenhouse cultivation of cut roses that is targeted at different varieties and takes into account the interaction of multiple environmental factors. Farmers and large production cooperatives or enterprises still guess the possible flowering date based on past experience. When encountering special weather changes, they cannot make timely adjustments to the environmental control settings. Therefore, there will be large differences in the guessed flowering period, and it is impossible to accurately predict the quality of cut flowers, resulting in losses in time, energy consumption and operations.

[0042] In addition to the supply of fixed order customers, the supply auction market, import and export market and special holiday market are also very important. These markets can often bring considerable orders and revenue, but they are also volatile. At present, producers do not have a good response strategy for this volatile supply, because the corresponding production plan formulated for changes in supply and demand will not achieve the expected results in actual implementation due to improper environmental control or untimely adjustment. Building a model can achieve repeated updates of predictions and efficient adjustments of relevant environmental impact factors. At the same time, it can also assist in formulating environmental control setting plans based on special market sales targets.

[0043] In actual production in modern protected areas, the entire rose plant is divided into two parts: the nutrient branch and the flower branch. The nutrient branch does not bloom but only grows leaves to continuously supplement photosynthetic assimilates for the growth of the flower branch that forms the output value. Therefore, the nutrient branch is usually regarded as the "source" of assimilates, while the flower branch is the "sink" of assimilates. The size relationship between the source and the sink, that is, the source-sink ratio, will affect the distribution of assimilates, and ultimately affect the yield and quality of the flower branch. In daily plant management, pruning is a means to control this ratio. It is often controlled by pruning and removing buds shortly after germination, and remains stable throughout the production. At present, the corresponding flowering period and quality prediction model algorithm established for different source-sink ratio scenarios has not been developed, and the survey of the source-sink quantity is not involved when formulating the production environment control plan. This item can cause errors in the prediction results and deviations between the formulated plan and the actual results. Supplementing the common source-sink ratio scenarios of different varieties in production can make the model closer to the actual situation and more accurate in actual application by combining the production density and source-sink quantity of field surveys.

[0044] An embodiment of the present application provides a method for constructing a flowering period and quality prediction model for the cut rose "Peach Snow Mountain". The model algorithm also supports reverse deduction to formulate a focused environmental condition control plan. Figure 1 and Figure 2 , Figure 1 A flowchart of a method for constructing a flowering period and quality prediction model for the cut rose "Peach Snow Mountain" provided in at least one embodiment of the present application, Figure 2A logic block diagram of a model construction method for constructing a flowering period and quality prediction model for a cut rose "Peach Snow Mountain" provided in at least one embodiment of the present application, Figure 2 The plant data obtained in (including stem diameter, stem length, stem dry weight, leaf area, leaf dry weight, bud height, bud dry weight, flowering period) are used for training the flowering period and quality prediction model. It should be understood that the method may also include additional boxes not shown and / or may omit the boxes shown, and the scope of the present application is not limited in this respect.

[0045] At step 101, data information at two important locations: the top of the vertical flower branch and the surface of the nutrient branch, is collected through sensors with the Internet of Things (for example, once every 30 minutes) from the internal environment data of the greenhouse where the cut rose "Peach Snow Mountain" is planted.

[0046] In some embodiments, the internal environment data acquired by the sensor is data information of the top of the vertical flower branch and the surface of the nutrient branch.

[0047] In some embodiments, the internal environment data includes at least one of photosynthetically active radiation intensity (400-700 nm, μmol m-2 s-1), temperature (°C), relative humidity (%), and carbon dioxide concentration (ppm).

[0048] At step 102, the internal environment data collected by the sensor is further subdivided and processed into a plurality of environmental factors to obtain input values.

[0049] In some embodiments, the input value includes at least one of the average daytime temperature, the average nighttime temperature, the average daytime humidity, the average nighttime humidity, the photosynthetically active radiation intensity, the carbon dioxide concentration, the relative radiant product and the radiant product.

[0050] In step 103, several common source-sink ratio processings are performed on the cut rose "Peach Snow Mountain" to obtain an output value.

[0051] In some embodiments, the source-sink ratio processing includes: obtaining 3 flower branches and 2 LAI (leaf area index) nutrient branch and leaf curtains, 3 flower branches and 4 LAI (leaf area index) nutrient branch and leaf curtains, 5 flower branches and 2 LAI (leaf area index) nutrient branch and leaf curtains, and 5 flower branches and 4 LAI (leaf area index) nutrient branch and leaf curtains, recording the dates of each growth period during the growth of the flower branches, including the germination period, bud period, and harvest period, and measuring the appearance quality at the time of harvest to obtain the output value, and the output value specifically includes: at least one of the flower branch length, flower branch thickness, flower bud height, leaf area and stem, leaf and flower dry weight.

[0052] At step 104 , multiple data sets including different input values ​​and corresponding output values ​​are generated by performing independent and random multiple sampling (bootstrap sampling) on ​​each of the output values ​​and all the input values ​​with replacement.

[0053] At step 105, a random forest algorithm (see formula 1-3) is used to obtain input values, ie, feature values, that have a high correlation with each output value, ie, predicted value.

[0054] In some embodiments, obtaining the characteristic value by using the random forest algorithm specifically includes:

[0055] The random forest algorithm quantifies the importance of features in two ways:

[0056] 1. Contribution of split points to impurity

[0057] For classification problems, feature importance, that is, the influence of the environmental factor, is calculated by the sum of the reductions in impurity caused by the feature in each node split of the decision tree. Gini impurity ( ) is reduced as follows:

[0058] (1)

[0059] in, is the impurity of the parent node, and is the impurity of the left and right nodes after the split, , , are the left and right child nodes and the total number of samples respectively.

[0060] feature The importance of is the sum of the impurity reduction of the split point based on this feature in all trees (see formula 2):

[0061] (2)

[0062] Among them, T is the index number of the decision tree, and s is the position of the split point.

[0063] 2. Permutation Importance

[0064] In order to further verify the impact of features, the permutation importance evaluation method is used. The core idea is to randomly disrupt the value of a single feature and observe the changes in model performance. The formula is as follows:

[0065] (3)

[0066] in, is the original performance of the model, is the performance of the model after the b-th permutation, and B is the number of permutation experiments.

[0067] At step 106, the relationship formula between each output value and its related characteristic value is further solved by polynomial fitting (see formula 4), so as to obtain the flowering period and quality prediction model.

[0068] Specifically include: Initialize the polynomial regression prediction model, assume that there is growth prediction features and merge all the features into one vector, i.e. , and at the same time set the initial parameter vector according to the number of features ,vector The initial values ​​of all parameter values ​​in are random values. The parameter vector is trained and adjusted according to the real data of the prediction target and feature value. The initial polynomial regression prediction model is:

[0069] (4)

[0070] In the formula, : Prediction value, i.e. output value, including flower branch thickness (cm), flower branch length (cm), leaf area (cm 2 ), bud height (cm), stem weight (g), flower dry weight (g);

[0071] : the i-th eigenvalue;

[0072] w i : The parameter corresponding to the i-th feature associated with the predicted target, w0 is the intercept term;

[0073] m: the power of the eigenvalue.

[0074] In some embodiments, if the degree of the fitting polynomial is too high, it is easy to overfit and produce abnormal fitting coefficients, then w is obtained by ridge regression. i Regularization is performed, that is, w i After differentiation, the result approaches 0.

[0075] The specific approach is: first, set the loss function for the algorithm, adjust the parameters according to the principle of minimum loss value, so as to improve the degree of fitting. The loss function is (see formula 5); then add a penalty term to the original polynomial function to obtain the formula of ridge regression (see formula 6)

[0076] (5)

[0077] (6)

[0078] In the formula, is the loss function, is the true value of growth, About The minimum value of To calculate the scalar and The square of the difference.

[0079] Through fitting, the relationship formula between all input characteristic values ​​(Table 1) and output prediction values ​​(Table 2) under common source-to-repository ratio settings is obtained (see Formula 10-35).

[0080]

[0081] (7)

[0082]

[0083] (9)

[0084] Where, d represents the dth day of the experiment, RTE(T) represents the physiological thermal effect when the average temperature is T, and the value range is 0-1; T b T is the lower limit temperature of development. When the temperature is lower than this, the development stops. For roses, it is 8℃. m The upper temperature limit of development, if the temperature exceeds this, the development will stop. For rose, it is 37℃; T ob The lower limit of the optimum temperature for development is 22℃ for roses. ou The upper limit of the optimum temperature for development is 25℃ for roses. The cumulative radiant heat product (TEP) is the daily accumulation of the relative radiant heat product RTEP, that is, the cumulative radiant heat product on the dth day is the accumulation of the relative radiant heat product of the day and the previous d-1 day.

[0085]

[0086] For all treatments:

[0087] (10)

[0088] (11)

[0089] Under the condition of retaining 5 flowering branches and a vegetative branch and leaf curtain with a leaf area index of 2:

[0090] (12)

[0091] (13)

[0092] (14)

[0093] (15)

[0094] (16)

[0095] (17)

[0096] Under the condition of retaining 3 flowering branches and a vegetative branch and leaf curtain with a leaf area index of 2:

[0097] (18)

[0098] (19)

[0099] (20)

[0100] (twenty one)

[0101] (twenty two)

[0102] (twenty three)

[0103] Under the condition of retaining 5 flowering branches and a vegetative branch curtain with a leaf area index of 4:

[0104] (twenty four)

[0105] (25)

[0106] (26)

[0107] (27)

[0108] (28)

[0109] (29)

[0110] Under the condition of retaining 3 flowering branches and a vegetative branch curtain with a leaf area index of 4:

[0111] (30)

[0112] (31)

[0113] (32)

[0114] (33)

[0115] (34)

[0116] (35)

[0117] In some embodiments, after obtaining the flowering period and quality prediction model, the method further includes:

[0118] Ensure that the cut rose "Peach Snow Mountain" in the greenhouse is planted in the same planting cycle with a spacing of 15 cm, a row spacing of 15 cm, a substrate trough spacing of 80 cm, and a leaf area index of 2 or 4. Obtain greenhouse internal environmental data at certain time intervals, and further segment and process them to obtain characteristic values. Substitute the processed characteristic values ​​into the flowering period and quality prediction model to obtain daily prediction values, i.e., output values.

[0119] Specifically, in actual use, if the user wants to use the flowering period and quality prediction model to repeatedly update the flowering period and quality prediction, the following prerequisites and information are required before use:

[0120] (1) The environmental sensors in the planting area of ​​this variety have a reasonable measurement range and take into account the height position of the flower branches and nutrient branches. They can collect all the internal environmental data of the greenhouse (temperature, relative humidity, photosynthetically active radiation intensity, carbon dioxide concentration) at a certain time interval: 15 minutes, 30 minutes or 1 hour, and can export the data through the client to facilitate data analysis and calculation;

[0121] (2) A five-point sampling method is used to conduct field surveys in the planting area of ​​the variety to understand the actual field spacing between plants, the spacing between soilless cultivation substrate troughs, the number of buds (retained buds) on each plant, and the date of budding. Ensure that within the same planting cycle, the planting spacing is 15 cm, the row spacing is 15 cm, and the spacing between substrate troughs is 80 cm. The number of buds must be consistent with the several scenarios that the model can provide for prediction: 3 (branches) or 5 (branches). The nutritional branches and leaves of the investigated plants need to be ( )Down( ) Photosynthetically active radiation intensity measurement, , and then put them into the Lambert-Beer formula (see formula 34) to calculate the leaf area index. Using the model for prediction requires ensuring that the leaf area index of the vegetative branches and leaves is consistent with the scenario that the model can provide predictions for: If the number of flower branches or the leaf area index of vegetative branches and leaves do not meet the conditions for the application of the model (the source and sink conditions do not meet), they need to be controlled through pruning;

[0122] (36)

[0123] (3) Starting from the germination day, the internal environment data is exported every day, and the internal environment data from 0:00 to 24:00 on the same day is further subdivided and processed according to the corresponding source library conditions and the characteristic value requirements involved in the content to be predicted, such as calculating the daytime average value and the nighttime maximum value;

[0124] The user can substitute the processed feature values ​​into the flowering period and quality prediction model under the corresponding source and database conditions to obtain the daily prediction value, that is, the output value. In addition, the input value obtained every day can be input into the flowering period and quality prediction model to obtain the corresponding output value and then update the output value obtained on the previous day.

[0125] In some embodiments, after obtaining the prediction results of flowering period and quality, the method further comprises:

[0126] Obtain characteristic values ​​for each predicted value under corresponding source-storage conditions, and perform environmental control settings based on the characteristic values ​​and actual production conditions.

[0127] Specifically, if auxiliary environmental control settings are required, you need to:

[0128] a. Specify the specific values ​​of flowering period and each output value;

[0129] b. The environmental sensors in the planting area of ​​this variety have a reasonable measurement range and take into account the height position of the flower branches and nutrient branches. They can collect all environmental data (temperature, relative humidity, photosynthetically active radiation intensity, carbon dioxide concentration) at a certain time interval: 15 minutes, 30 minutes or 1 hour, and can export data through the client to facilitate data analysis and calculation;

[0130] c. Use the five-point sampling method to conduct field surveys in the planting area of ​​this variety to understand the actual field spacing between plants and rows, the spacing between soilless cultivation substrate troughs, the number of buds (retained buds) on each plant, and the date of budding. Ensure that within the same cropping cycle, the spacing between plants is 15 cm, the spacing between rows is 15 cm, and the spacing between substrate troughs is 80 cm. The number of buds must be consistent with the several scenarios that the model can provide for prediction: 3 (branches) or 5 (branches). The nutritional branches and leaves of the investigated plants need to be ( )Down( ) Photosynthetically active radiation intensity measurement, , and then put them into the Lambert-Beer formula (see formula 34) to calculate the leaf area index. Using the model for prediction requires ensuring that the leaf area index of the vegetative branches and leaves is consistent with the scenario that the model can provide predictions for: If the number of flower branches or the leaf area index of vegetative branches and leaves do not meet the conditions for the application of the model (the source and sink conditions do not meet), they need to be controlled through pruning;

[0131] The characteristic values ​​of each predicted value under the corresponding source and storage conditions, namely the environmental factors and setting values, are solved through the formula, and the environmental control settings can be made based on the actual production situation.

[0132] The final fit, that is, the prediction accuracy, is based on the mean square error (MSE) between the predicted value and the measured value, and the formula is:

[0133] (37)

[0134] in, is the observed value, is the predicted value, and n is the number of predictions. The smaller the MSE, the better the prediction effect.

[0135] By comparing with the actual measured values ​​of the validation group, the MSE of the prediction results under various source-to-sink ratio conditions is shown in Table 3.

[0136]

[0137] Table 3:

[0138]

[0139] Figure 3 A block diagram of a module for constructing a flowering period and quality prediction model for a cut rose "Peach Snow Mountain" provided in at least one embodiment of the present application. The device includes:

[0140] A memory 201; and a processor 202 connected to the memory 201, wherein the processor 202 is configured to: collect internal environmental data of a greenhouse where a cut rose "Peach Snow Mountain" is planted through a sensor with an Internet of Things;

[0141] Further subdividing and processing the internal environment data collected by the sensor into multiple environmental factors to obtain input values;

[0142] The source-sink ratio of the cut rose "Peach Snow Mountain" was processed to obtain the output value;

[0143] Generate multiple data sets including different input values ​​and corresponding output values ​​by sampling each of the output values ​​and all the input values ​​independently and randomly with replacement multiple times;

[0144] The random forest algorithm is used to obtain input values ​​that have a high correlation with each output value, namely, the eigenvalues;

[0145] The relationship formula between each output value and its related characteristic value is solved by polynomial fitting, so as to obtain the flowering period and quality prediction model.

[0146] In some embodiments, the processor 202 is further configured to: the internal environment data acquired by the sensor is data information of the top of the vertical flower branch and the surface of the nutrient branch;

[0147] The internal environment data includes at least one of photosynthetically active radiation intensity, temperature, relative humidity and carbon dioxide concentration.

[0148] In some embodiments, the processor 202 is further configured such that: the input value includes at least one of the average daytime temperature, the average nighttime temperature, the average daytime humidity, the average nighttime humidity, the photosynthetically active radiation intensity, the carbon dioxide concentration, the relative radiant product and the radiant product.

[0149] In some embodiments, the processor 202 is further configured to: the source-sink ratio processing includes: obtaining 3 flower branches and 2 leaf area index nutrient branch and leaf curtain, 3 flower branches and 4 leaf area index nutrient branch and leaf curtain, 5 flower branches and 2 leaf area index nutrient branch and leaf curtain, 5 flower branches and 4 leaf area index nutrient branch and leaf curtain, recording the dates of each growth period during the growth of the flower branches, including the germination period, budding period, and harvesting period, measuring the appearance quality at harvest time, and obtaining the output value, the output value specifically includes: at least one of the flower branch length, flower branch thickness, flower bud height, leaf area and stem, leaf and flower dry weight.

[0150] In some embodiments, the processor 202 is further configured to: obtain the feature value by using the random forest algorithm, specifically including:

[0151] The influence of the environmental factor is calculated by summing the reduction in impurity caused by the feature in the node splitting of each decision tree;

[0152] The permutation importance assessment method was used to further verify the impact of features.

[0153] In some embodiments, the processor 202 is further configured to: solve the relationship formula between each output value and its related characteristic value by polynomial fitting to obtain the flowering period and quality prediction model, specifically including:

[0154] Initialize the polynomial regression prediction model, assuming that there is growth prediction features and merge all the features into one vector, i.e. , and at the same time set the initial parameter vector according to the number of features ,vector The initial values ​​of all parameter values ​​in are random values. The parameter vector is trained and adjusted according to the real data of the prediction target and feature value. The initial polynomial regression prediction model is:

[0155] ,

[0156] In the formula, Represents the predicted value, i.e., the output value, including flower branch thickness, flower branch length, leaf area, bud height, stem weight, and flower dry weight;

[0157] represents the i-th eigenvalue;

[0158] w i It represents the parameter corresponding to the i-th feature associated with the predicted target, and w0 is the intercept term;

[0159] m represents the power of the eigenvalue.

[0160] In some embodiments, the processor 202 is further configured to: if the degree of the fitted polynomial is too high, perform ridge regression on w i Regularization is performed, that is, w i After differentiation, the result approaches 0.

[0161] In some embodiments, the processor 202 is further configured to: after obtaining the flowering period and quality prediction model, further include:

[0162] Ensure that the cut rose "Peach Snow Mountain" in the greenhouse is planted in the same planting cycle with a spacing of 15 cm, a row spacing of 15 cm, a substrate trough spacing of 80 cm, and a leaf area index of 2 or 4, obtain the internal environment data of the greenhouse at a certain time interval, and further subdivide and process it to obtain characteristic values, and substitute the processed characteristic values ​​into the flowering period and quality prediction model to obtain daily prediction values, that is, output values;

[0163] After obtaining the prediction results of flowering period and quality, it also includes:

[0164] Obtain characteristic values ​​for each predicted value under corresponding source-storage conditions, and perform environmental control settings based on the characteristic values ​​and actual production conditions.

[0165] The specific implementation method is referred to the aforementioned method embodiment and will not be described in detail here.

[0166] The present application may be a method, an apparatus, a system and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present application.

[0167] Computer readable storage medium can be a tangible device that can hold and store instructions used by an instruction execution device. Computer readable storage medium can be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination thereof. More specific examples (non-exhaustive list) of computer readable storage medium include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, for example, a punch card or a convex structure in a groove on which instructions are stored, and any suitable combination thereof. The computer readable storage medium used here is not interpreted as a transient signal itself, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagated by a waveguide or other transmission medium (for example, a light pulse by an optical fiber cable), or an electrical signal transmitted by a wire.

[0168] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0169] The computer program instructions for performing the operation of the present application may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. Computer-readable program instructions may be executed entirely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet). In some embodiments, an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), may be personalized by utilizing the state information of the computer-readable program instructions, and the electronic circuit may execute the computer-readable program instructions, thereby realizing various aspects of the present application.

[0170] Various aspects of the present application are described herein with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present application. It should be understood that each box in the flowchart and / or block diagram and the combination of boxes in the flowchart and / or block diagram can be implemented by computer-readable program instructions.

[0171] These computer-readable program instructions can be provided to a processing unit of a general-purpose computer, a special-purpose computer, or other programmable data processing device, thereby producing a machine, so that when these instructions are executed by the processing unit of the computer or other programmable data processing device, a device that implements the functions / actions specified in one or more boxes in the flowchart and / or block diagram is generated. These computer-readable program instructions can also be stored in a computer-readable storage medium, and these instructions cause the computer, programmable data processing device, and / or other equipment to work in a specific manner, so that the computer-readable medium storing the instructions includes a manufactured product, which includes instructions for implementing various aspects of the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0172] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operating steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to implement the functions / actions specified in one or more boxes in the flowchart and / or block diagram.

[0173] The flow chart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to multiple embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a part of a module, program segment or instruction, and a part of the module, program segment or instruction includes one or more executable instructions for realizing the logical function of the specification. In some alternative implementations, the function marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two continuous square boxes can actually be executed substantially in parallel, and they can also be executed in the opposite order sometimes, depending on the function involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be realized by a dedicated hardware-based system that performs the function or action of the specification, or can be realized by a combination of special-purpose hardware and computer instructions.

[0174] Note that, unless otherwise directly stated, all features disclosed in this specification (including any attached claims, abstracts and drawings) may be replaced by alternative features for achieving the same, equivalent or similar purposes. Therefore, unless otherwise explicitly stated, each feature disclosed is only an example of a group of equivalent or similar features. Where used, further, preferably, further and more preferably are simple beginnings for elaborating another embodiment based on the aforementioned embodiment, and the content of further, preferably, further or more preferably followed by the aforementioned embodiment is combined with the complete composition of another embodiment. Several further, preferably, further or more preferably settings following the same embodiment can be arbitrarily combined to form another embodiment.

[0175] Although the present application has been described in detail above with general descriptions and specific embodiments, it is obvious to those skilled in the art that some modifications or improvements may be made to the present application. Therefore, these modifications or improvements made without departing from the spirit of the present application all fall within the scope of protection claimed in the present application.

Claims

1. A method for constructing a flowering period and quality prediction model for cut rose "Peach Snow Mountain", characterized in that: The following steps are involved: Collecting internal environmental data of the greenhouse where the cut rose "Peach Snow Mountain" is planted through sensors with the Internet of Things; Further subdividing and processing the internal environment data collected by the sensor into multiple environmental factors to obtain input values; Perform source-sink ratio processing on the cut rose "Peach Snow Mountain" to obtain the output value; Generate multiple data sets including different input values ​​and corresponding output values ​​by sampling each of the output values ​​and all the input values ​​independently and randomly with replacement multiple times; The random forest algorithm is used to obtain input values ​​that have a high correlation with each output value, namely, the eigenvalues; The relationship formula between each output value and its related characteristic value is solved by polynomial fitting, so as to obtain the flowering period and quality prediction model.

2. The method for constructing the cut flower rose "Peach Snow Mountain" florescence and quality prediction model according to claim 1, characterized in that, The internal environment data acquired by the sensor is data information of the top of the vertical flower branch and the surface of the nutrient branch; The internal environment data includes at least one of photosynthetically active radiation intensity, temperature, relative humidity and carbon dioxide concentration.

3. The method for constructing the cut flower rose "Peach Snow Mountain" florescence and quality prediction model according to claim 1, characterized in that, The input values ​​include at least one of the daytime average temperature, the nighttime average temperature, the daytime average humidity, the nighttime average humidity, the photosynthetically active radiation intensity, the carbon dioxide concentration, the relative radiant product and the radiant product.

4. The method for constructing the cut flower rose "Peach Snow Mountain" florescence and quality prediction model according to claim 1, characterized in that, The source-sink ratio processing includes: obtaining 3 flower branches and 2 leaf area index vegetative branch and leaf curtains, 3 flower branches and 4 leaf area index vegetative branch and leaf curtains, 5 flower branches and 2 leaf area index vegetative branch and leaf curtains, and 5 flower branches and 4 leaf area index vegetative branch and leaf curtains, recording the dates of each growth period during the growth of the flower branches, including the germination period, the budding period, and the harvesting period, measuring the appearance quality at the time of harvest, and obtaining the output value, which specifically includes: at least one of the flower branch length, flower branch thickness, flower bud height, leaf area, and stem, leaf, and flower dry weight.

5. The method for constructing the cut flower rose "Peach Snow Mountain" florescence and quality prediction model according to claim 1, characterized in that, The use of the random forest algorithm to obtain the characteristic value specifically includes: The influence of the environmental factor is calculated by summing the reduction in impurity caused by the feature in the node splitting of each decision tree; The permutation importance assessment method was used to further verify the impact of features.

6. The method for constructing the cut flower rose "Peach Snow Mountain" florescence and quality prediction model according to claim 1, characterized in that, The polynomial fitting is used to solve the relationship formula between each output value and its related characteristic value, so as to obtain the flowering period and quality prediction model, which specifically includes: Initialize the polynomial regression prediction model, assuming that there is growth prediction features and merge all the features into one vector, i.e. , and at the same time set the initial parameter vector according to the number of features ,vector The initial values ​​of all parameter values ​​in are random values. The parameter vector is adjusted through training based on the real data of the prediction target and feature value. The initial polynomial regression prediction model is: ， In the formula, Represents the predicted value, i.e., the output value, including flower branch thickness, flower branch length, leaf area, bud height, stem weight, and flower dry weight; represents the i-th eigenvalue; w i It represents the parameter corresponding to the i-th feature associated with the predicted target, and w0 is the intercept term; m represents the power of the eigenvalue.

7. The method for constructing the cut flower rose "Peach Snow Mountain" florescence and quality prediction model according to claim 6, characterized in that, If the degree of the fitted polynomial is too high, then w is i Regularization is performed, that is, w i After differentiation, the result approaches 0.

8. The method for constructing the cut flower rose "Peach Snow Mountain" florescence and quality prediction model according to claim 1, characterized in that, After obtaining the flowering period and quality prediction model, the method further includes: Ensure that the cut rose "Peach Snow Mountain" in the greenhouse is planted in the same planting cycle with a spacing of 15 cm, a row spacing of 15 cm, a substrate trough spacing of 80 cm, and a leaf area index of 2 or 4. Obtain greenhouse internal environmental data at certain time intervals, further subdivide and process them, obtain characteristic values, and substitute the processed characteristic values ​​into the flowering period and quality prediction model to obtain daily prediction values, i.e., output values; After obtaining the prediction results of flowering period and quality, it also includes: Obtain characteristic values ​​for each predicted value under corresponding source-storage conditions, and perform environmental control settings based on the characteristic values ​​and actual production conditions.

9. A device for constructing a flowering period and quality prediction model for the cut rose "Peach Snow Mountain", characterized in that: include: Memory; as well as A processor connected to the memory, the processor being configured to perform the steps of the method according to any one of claims 1 to 8.

10. A computer storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a machine, the steps of the method according to any one of claims 1 to 8 are implemented.

Images