Parameter adjustment method and device for crop simulation model, processor and storage medium
By screening and adjusting the genetic parameter combination of crop simulation models, the problem of low parameter adjustment efficiency is solved, efficient and accurate parameter optimization is achieved, and the accuracy of the simulation model is improved.
Patent Information
- Application Number
- CN202510314265.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2024-04-03
- Filing Date
- 2025-03-17
- Publication Date
- 2025-07-18
AI Technical Summary
The existing crop simulation models are not efficient when adjusting parameters and cannot adapt to the differences in different crop breeding methods, resulting in a long parameter adjustment time.
By obtaining crop planting data and initial genetic parameter groups, the accuracy of simulation results is determined, and the genetic parameter groups are screened layer by layer, including coarse adjustment and fine adjustment, adjusting the parameter value range, optimizing the genetic parameter combination, and improving the accuracy of the simulation model.
The parameter adjustment time is shortened, the parameter adjustment efficiency and accuracy is improved, the trial and error range of genetic parameters is ensured, and the loss of important parameter values is avoided, and the accuracy of simulation results is improved.
Smart Images
Figure CN120337708A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of agricultural information technology, and particularly relates to a method and device for adjusting parameters of a crop simulation model, a processor, and a storage medium. Background Art
[0002] A crop simulation model, also known as a crop growth simulation model, is a tool that uses mathematical models and computer technology to simulate and predict processes related to crop growth, development, yield, and quality. By modeling and simulating factors such as the crop growth environment, soil conditions, meteorological conditions, and crop physiological and biochemical processes, it can help agricultural scientists, farmers, and government decision-makers make decisions in aspects such as crop production management, climate change adaptation, and food security assessment. Crop simulation models are of great significance for crop yield and the sustainable development of agricultural production. In the prior art, when adjusting parameters of a crop simulation model, the value range of genetic parameters is usually fixed. However, with different crop breeding methods, the genetic parameters of the crop simulation model for the same variety of crops will also vary. Adjusting parameters according to the genetic parameters with a fixed value range takes a long time and the parameter adjustment efficiency is not high. Therefore, there is a problem of low parameter adjustment efficiency for crop simulation models in the prior art. Summary of the Invention
[0003] The purpose of the embodiments of this application is to provide a method and device for adjusting parameters of a crop simulation model, a processor, and a storage medium, so as to solve the problem of low parameter adjustment efficiency for crop simulation models in the prior art.
[0004] To achieve the above purpose, the first aspect of this application provides a method for adjusting parameters of a crop simulation model. The parameter adjustment method includes:
[0005] Obtain the input parameters and initial model parameters of the crop simulation model, where the input parameters include crop planting data, and the initial model parameters include multiple initial genetic parameter groups, and the value of each genetic parameter in each initial genetic parameter group does not exceed the initial parameter value range of each genetic parameter;
[0006] Determine the first simulation result accuracy of the crop simulation model under each initial genetic parameter group according to the crop planting data and the multiple initial genetic parameter groups;
[0007] Screen the multiple initial genetic parameter groups according to the first simulation result accuracy to obtain multiple roughly adjusted genetic parameter groups, where the roughly adjusted genetic parameter groups are the initial genetic parameter groups in the multiple initial genetic parameter groups whose first simulation result accuracy reaches the first preset accuracy;
[0008] Determine the maximum value and minimum value of each genetic parameter in the multiple roughly adjusted genetic parameter groups to obtain the adjustment parameter value range of each genetic parameter;
[0009] Determine multiple fine-tuned genetic parameter groups according to the adjustment parameter value ranges of the respective genetic parameters;
[0010] According to the crop planting data and the multiple fine-tuned genetic parameter groups, determine the second simulation result accuracy of the crop simulation model under each fine-tuned genetic parameter group;
[0011] Determine the target genetic parameter group among the multiple fine-tuned genetic parameter groups according to the second simulation result accuracy, so as to obtain the final model parameters of the crop simulation model, where the target genetic parameter group is the fine-tuned genetic parameter group in which the second simulation result accuracy among the multiple fine-tuned genetic parameter groups reaches a second preset accuracy, and the second preset accuracy is greater than the first preset accuracy.
[0012] In the embodiments of the present application, determining the first simulation result accuracy of the crop simulation model under each initial genetic parameter group according to the crop planting data and the multiple initial genetic parameter groups includes: inputting the crop planting data into the crop simulation models corresponding to the respective initial genetic parameter groups to obtain the first simulation results output by the crop simulation models corresponding to the respective initial genetic parameter groups; determining the first simulation result accuracy of the crop simulation model under each initial genetic parameter group according to the first simulation results and the true results corresponding to the pre-obtained crop planting data; determining the second simulation result accuracy of the crop simulation model under each fine-tuned genetic parameter group according to the crop planting data and the multiple fine-tuned genetic parameter groups includes: inputting the crop planting data into the crop simulation models corresponding to the respective fine-tuned genetic parameter groups to obtain the second simulation results output by the crop simulation models corresponding to the respective fine-tuned genetic parameter groups; determining the second simulation result accuracy of the crop simulation model under each fine-tuned genetic parameter group according to the second simulation results and the true results corresponding to the pre-obtained crop planting data.
[0013] In the embodiments of the present application, the first simulation result accuracy is determined according to the deviation between the first simulation result and the true result; the second simulation result accuracy is determined according to the deviation between the second simulation result and the true result.
[0014] In the embodiments of the present application, the number of the first simulation results is multiple, including multiple first simulation results output by the crop simulation models corresponding to the respective initial genetic parameter groups after multiple groups of crop planting data are respectively input into the crop simulation models corresponding to the respective initial genetic parameter groups, and the first simulation result accuracy is determined according to the deviations between the multiple first simulation results and the true results corresponding to the respective first simulation results; the number of the second simulation results is multiple, including multiple second simulation results output by the crop simulation models corresponding to the respective fine-tuned genetic parameter groups after multiple groups of crop planting data are respectively input into the crop simulation models corresponding to the respective fine-tuned genetic parameter groups, and the second simulation result accuracy is determined according to the deviations between the multiple second simulation results and the true results corresponding to the respective second simulation results.
[0015] In the embodiments of the present application, according to the adjustment parameter value ranges of the respective genetic parameters, a plurality of fine-tuned genetic parameter groups are determined, including: according to the adjustment parameter value ranges of the respective genetic parameters, a plurality of fine-tuned genetic parameter groups are determined by using the exhaustive method.
[0016] In the embodiments of the present application, the output parameters of the crop simulation model include growth period parameters and / or yield parameters.
[0017] In the embodiments of the present application, the number of the plurality of initial genetic parameter groups is within a preset number range.
[0018] The second aspect of the present application provides a processor configured to execute the parameter adjustment method for the crop simulation model according to the above.
[0019] The third aspect of the present application provides a parameter adjustment device for a crop simulation model, including: the processor according to the above.
[0020] The fourth aspect of the present application provides a machine-readable storage medium, on which instructions are stored, and the instructions are used to cause a machine to execute the parameter adjustment method for the crop simulation model according to the above.
[0021] In the embodiments of the present application, by using the crop planting data and a plurality of initial genetic parameter groups determined based on the initial parameter value ranges of the respective genetic parameters, the first simulation result accuracy of the crop simulation model under each initial genetic parameter group is determined, and the plurality of initial genetic parameter groups are screened according to the first simulation result accuracy to obtain a plurality of roughly adjusted genetic parameter groups. Based on the maximum and minimum values of the respective genetic parameters in the plurality of roughly adjusted genetic parameter groups, the adjustment parameter value ranges of the respective genetic parameters are determined. Furthermore, according to the adjustment parameter value ranges of the respective genetic parameters, a plurality of fine-tuned genetic parameter groups are determined, and according to the second simulation result accuracy of the crop simulation model under each fine-tuned genetic parameter group, a target genetic parameter group in the plurality of fine-tuned genetic parameter groups is determined to obtain the final model parameters of the crop simulation model. In the above technical solution, through layer-by-layer screening, the value ranges of the genetic parameters of the crop simulation model are changed, and the rough adjustment and fine adjustment of the genetic parameters of the crop simulation model can be realized, the parameter adjustment duration is shortened, and the parameter adjustment efficiency is improved. In addition, while ensuring that the value ranges of the genetic parameters are narrowed, the adjustment parameter value ranges of the respective genetic parameters are determined according to the maximum and minimum values of the respective genetic parameters in the plurality of roughly adjusted genetic parameter groups, ensuring a relatively large trial error range for the genetic parameters, not easily losing important parameter values, and avoiding losing the optimal solution, thereby being able to improve the parameter adjustment accuracy of the genetic parameters of the crop simulation model, and further improving the accuracy of the simulation results of the crop simulation model.
[0022] Other features and advantages of the embodiments of the present application will be described in detail in the subsequent specific implementation part. Description of the Drawings
[0023] The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the embodiments of the present application, but do not constitute a limitation to the embodiments of the present application. In the accompanying drawings:
[0024] Figure 1 A schematic flowchart of a method for adjusting parameters of a crop simulation model according to an embodiment of the present application is schematically shown. Specific Embodiments
[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present application. It should be understood that the specific embodiments described herein are only used to illustrate and explain the embodiments of the present application and do not limit the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts fall within the scope of protection of the present application.
[0026] It should be noted that the acquisition, transmission, storage, use, processing, etc. of data in the technical solutions of the present application all comply with the relevant regulations of national laws and regulations. In the embodiments of the present application, certain industry-existing solutions such as software, components, models, etc. may be mentioned. They should be regarded as exemplary. The purpose is only to illustrate the feasibility in the implementation of the technical solutions of the present application, but it does not mean that the applicant has already or necessarily used this solution.
[0027] It should be noted that if there are directional indications (such as up, down, left, right, front, back...) involved in the embodiments of the present application, then the directional indications are only used to explain the relative positional relationship and movement conditions between components in a specific posture (as shown in the accompanying drawings). If this specific posture changes, then the directional indications will also change accordingly.
[0028] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present application, then the descriptions of "first", "second", etc. are only for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one such feature. In addition, the technical solutions between various embodiments can be combined with each other, but it must be based on the ability of those of ordinary skill in the art to implement. When the combination of technical solutions results in contradictions or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by the present application.
[0029] Figure 1The figure schematically shows a flowchart of a method for adjusting parameters of a crop simulation model according to an embodiment of the present application. As Figure 1 shown, an embodiment of the present application provides a method for adjusting parameters of a crop simulation model. Taking the application of this method to a processor as an example, this method may include the following steps.
[0030] Step S102, obtain the input parameters and initial model parameters of the crop simulation model. Among them, the input parameters include crop planting data, and the initial model parameters include multiple initial genetic parameter groups. The values of each genetic parameter in each initial genetic parameter group do not exceed the initial parameter value range of each genetic parameter.
[0031] It can be understood that the crop simulation model, also known as the crop growth simulation model, is a type of agricultural mathematical model that quantitatively describes the dynamic processes of crop growth, development, grain formation, and yield based on meteorological conditions, soil conditions, and crop cultivation management measures. For example, it can be the open-source DSSAT model. The input of the crop simulation model can include crop planting data, which is the relevant data of the crop during the planting process, and specifically can include but is not limited to data such as basic data, fertilization data, irrigation data, soil data, observation data, and meteorological data. The crop can be, for example, rice, wheat, etc. The genetic parameter is a parameter of the crop simulation model, specifically referring to the parameter used to describe the growth and development traits of the crop in the crop simulation model, and its value is usually multiple. Different crop simulation models have different genetic parameters. The more accurate the parameter values of the genetic parameters, the better the simulation effect of the crop simulation model on crop growth. The initial model parameters are multiple groups of parameters of the crop simulation model preliminarily determined based on the initial parameter value range of each genetic parameter, and can include multiple initial genetic parameter groups. The initial genetic parameter group is a genetic parameter group containing multiple genetic parameters. The values of each genetic parameter in the initial genetic parameter group do not exceed the initial parameter value range of each genetic parameter. The initial parameter value range is the initial value range of the genetic parameter set in advance, and its lower threshold is the initialization starting value of the genetic parameter, and the upper threshold is the initialization termination value of the genetic parameter.
[0032] Specifically, the processor can obtain the input parameters of the crop simulation model determined in advance and the initial model parameters of the crop simulation model. The input parameters can include crop planting data, and the initial model parameters can include multiple initial genetic parameter groups. The multiple initial genetic parameter groups can be randomly selected and determined according to the initial parameter value range of each genetic parameter. At least one genetic parameter value in each initial genetic parameter group is different.
[0033] Step S104, determine the first simulation result accuracy of the crop simulation model under each initial genetic parameter group according to the crop planting data and the multiple initial genetic parameter groups.
[0034] It can be understood that the accuracy of the first simulation result is the accuracy of the simulation results output by the crop simulation model corresponding to each initial genetic parameter group. The parameters related to the simulation results can be multiple or one. For example, the accuracy of the first simulation result can be specifically determined according to the deviation between the simulation result and the real result. The greater the accuracy of the first simulation result, the more accurate the simulation result of the crop simulation model is and the closer it is to the real data of the crop.
[0035] Specifically, the processor can use the crop planting data as the input of the crop simulation model and change the model parameters of the crop simulation model. That is to say, the model parameters of the crop simulation model are set to each initial genetic parameter group, so that the accuracy of the first simulation result of the crop simulation model under each initial genetic parameter group can be obtained.
[0036] Step S106: Screen multiple initial genetic parameter groups according to the accuracy of the first simulation result to obtain multiple roughly adjusted genetic parameter groups, where the roughly adjusted genetic parameter group is the initial genetic parameter group in which the accuracy of the first simulation result reaches the first preset accuracy among the multiple initial genetic parameter groups.
[0037] It can be understood that the first preset accuracy is the accuracy threshold of the simulation result of the crop simulation model set in advance, and its value can take a value above medium. The roughly adjusted genetic parameter group is the genetic parameter group determined after roughly screening multiple initial genetic parameter groups, and its number is multiple. Specifically, it is the initial genetic parameter group in which the accuracy of the first simulation result is greater than or equal to the first preset accuracy.
[0038] Specifically, after the processor obtains the accuracy of the first simulation result corresponding to each initial genetic parameter group, it can screen multiple initial genetic parameter groups based on the first preset accuracy, so as to screen out multiple roughly adjusted genetic parameter groups in which the accuracy of the first simulation result reaches the first preset accuracy for subsequent further refined parameter adjustment.
[0039] Step S108: Determine the maximum and minimum values of each genetic parameter in the multiple roughly adjusted genetic parameter groups to obtain the adjustment parameter value range of each genetic parameter.
[0040] It can be understood that the adjustment parameter value range is the adjusted value range of each genetic parameter determined based on the multiple roughly adjusted genetic parameter groups. Specifically, the adjustment parameter value range of each genetic parameter can be determined according to the maximum and minimum values of each genetic parameter in the multiple roughly adjusted genetic parameter groups.
[0041] Specifically, the processor can determine the maximum and minimum values of each genetic parameter in multiple coarse-tuning genetic parameter groups to obtain the adjustment parameter value range of each genetic parameter. That is, the upper threshold of the adjustment parameter value range of each genetic parameter is the maximum value of each genetic parameter in multiple coarse-tuning genetic parameter groups, and the lower threshold is the minimum value of each genetic parameter in multiple coarse-tuning genetic parameter groups.
[0042] Step S110, determine multiple fine-tuning genetic parameter groups according to the adjustment parameter value range of each genetic parameter.
[0043] It can be understood that the fine-tuning genetic parameter groups are genetic parameter groups re-determined based on the adjustment parameter value range of each genetic parameter, and the number of them is multiple.
[0044] Specifically, the processor can determine multiple fine-tuning genetic parameter groups according to the adjustment parameter value range of each genetic parameter. From this, it can be seen that the adjustment parameter value range of each genetic parameter is further narrowed compared with the initial parameter value range of each genetic parameter. Since the adjustment parameter value range is determined based on the maximum and minimum values of each genetic parameter in multiple coarse-tuning genetic parameter groups, a larger trial-and-error range of parameters is ensured, important parameter values are not easily lost, and the optimal solution is not lost, thereby improving the parameter adjustment accuracy of the genetic parameters of the crop simulation model.
[0045] Step S112, determine the second simulation result accuracy of the crop simulation model under each fine-tuning genetic parameter group according to the crop planting data and multiple fine-tuning genetic parameter groups.
[0046] It can be understood that the second simulation result accuracy is the accuracy of the simulation results output by the crop simulation model corresponding to each fine-tuning genetic parameter group. The parameters related to the simulation results can be multiple or one. The second simulation result accuracy can be specifically determined according to the deviation between the simulation results and the real results, for example. The greater the second simulation result accuracy, the more accurate the simulation results of the crop simulation model are and the closer they are to the real data of the crop.
[0047] Specifically, the processor can use the crop planting data as the input of the crop simulation model and change the model parameters of the crop simulation model. That is, set the model parameters of the crop simulation model to each fine-tuning genetic parameter group, so as to obtain the second simulation result accuracy of the crop simulation model under each fine-tuning genetic parameter group.
[0048] Step S114, determine the target genetic parameter group in multiple fine-tuning genetic parameter groups according to the second simulation result accuracy to obtain the final model parameters of the crop simulation model, where the target genetic parameter group is the fine-tuning genetic parameter group in multiple fine-tuning genetic parameter groups whose second simulation result accuracy reaches the second preset accuracy, and the second preset accuracy is greater than the first preset accuracy.
[0049] It can be understood that the second preset accuracy is the accuracy threshold of the simulation results of the pre-set crop simulation model, and its value can be a relatively high value. Since the first preset accuracy is used to preliminarily screen multiple initial genetic parameter groups to obtain multiple roughly adjusted genetic parameter groups, and the second preset accuracy is used to finely screen multiple finely adjusted genetic parameter groups to determine the target genetic parameter group, the second preset accuracy is greater than the first preset accuracy. The target genetic parameter group is a combination of the ideal values of the genetic parameters of the crop simulation model finally determined after parameter adjustment, that is, the final model parameters of the crop simulation model. It is determined by screening multiple finely adjusted genetic parameter groups. Specifically, it is the finely adjusted genetic parameter group in which the second simulation result accuracy in multiple finely adjusted genetic parameter groups reaches the second preset accuracy.
[0050] Specifically, the processor can compare the second simulation result accuracy of the crop simulation model under each finely adjusted genetic parameter group, and determine the finely adjusted genetic parameter group with the highest second simulation result accuracy as the target genetic parameter group to obtain the final model parameters of the crop simulation model.
[0051] In the embodiment of the present application, based on multiple initial genetic parameter groups determined according to crop planting data and the initial parameter value ranges based on each genetic parameter, the first simulation result accuracy of the crop simulation model under each initial genetic parameter group is determined. Multiple initial genetic parameter groups are screened according to the first simulation result accuracy to obtain multiple roughly adjusted genetic parameter groups. Based on the maximum and minimum values of each genetic parameter in multiple roughly adjusted genetic parameter groups, the adjustment parameter value range of each genetic parameter is determined. Furthermore, multiple finely adjusted genetic parameter groups are determined according to the adjustment parameter value range of each genetic parameter, and the target genetic parameter group in multiple finely adjusted genetic parameter groups is determined according to the second simulation result accuracy of the crop simulation model under each finely adjusted genetic parameter group to obtain the final model parameters of the crop simulation model. The above technical solution can change the value range of the genetic parameters of the crop simulation model through layer-by-layer screening, realize the rough adjustment and fine adjustment of the genetic parameters of the crop simulation model, shorten the parameter adjustment time, and improve the parameter adjustment efficiency. In addition, while ensuring that the value range of the genetic parameters is narrowed, the adjustment parameter value range of each genetic parameter is determined according to the maximum and minimum values of each genetic parameter in multiple roughly adjusted genetic parameter groups, ensuring a large trial-and-error range for the genetic parameters, not easily losing important parameter values, and avoiding losing the optimal solution. Therefore, the parameter adjustment accuracy of the genetic parameters of the crop simulation model can be improved, and further the accuracy of the simulation results of the crop simulation model can be improved.
[0052] In one embodiment, according to crop planting data and multiple initial genetic parameter groups, determining the accuracy of the first simulation result of the crop simulation model under each initial genetic parameter group includes: inputting the crop planting data into the crop simulation model corresponding to each initial genetic parameter group to obtain the first simulation result output by the crop simulation model corresponding to each initial genetic parameter group; determining the accuracy of the first simulation result of the crop simulation model under each initial genetic parameter group according to the first simulation result and the true result corresponding to the pre-acquired crop planting data.
[0053] It can be understood that the first simulation result is the simulation result output by the crop simulation model corresponding to the initial genetic parameter group after inputting the crop planting data into the crop simulation model corresponding to the initial genetic parameter group.
[0054] Specifically, the processor can input the crop planting data into the crop simulation model corresponding to each initial genetic parameter group to obtain the first simulation result output by the crop simulation model corresponding to each initial genetic parameter group, and then determine the accuracy of the first simulation result of the crop simulation model under each initial genetic parameter group according to the first simulation result and the true result corresponding to the pre-acquired crop planting data.
[0055] In one embodiment, the accuracy of the first simulation result is determined according to the deviation between the first simulation result and the true result.
[0056] Specifically, the accuracy of the first simulation result can be the reciprocal of the deviation between the first simulation result and the true result. The greater the deviation, the smaller the accuracy of the first simulation result. The deviation can be determined by means such as absolute deviation and relative deviation.
[0057] In one embodiment, according to crop planting data and multiple fine-tuned genetic parameter groups, determining the accuracy of the second simulation result of the crop simulation model under each fine-tuned genetic parameter group includes: inputting the crop planting data into the crop simulation model corresponding to each fine-tuned genetic parameter group to obtain the second simulation result output by the crop simulation model corresponding to each fine-tuned genetic parameter group; determining the accuracy of the second simulation result of the crop simulation model under each fine-tuned genetic parameter group according to the second simulation result and the true result corresponding to the pre-acquired crop planting data.
[0058] It can be understood that the second simulation result is the simulation result output by the crop simulation model corresponding to the fine-tuned genetic parameter group after inputting the crop planting data into the crop simulation model corresponding to the fine-tuned genetic parameter group.
[0059] Specifically, the processor may input the crop planting data into the crop simulation models corresponding to each fine-tuned genetic parameter group to obtain the second simulation results output by the crop simulation models corresponding to each fine-tuned genetic parameter group, and then determine the accuracy of the second simulation results of the crop simulation models under each fine-tuned genetic parameter group according to the second simulation results and the true results corresponding to the pre-acquired crop planting data.
[0060] In one embodiment, the accuracy of the second simulation results is determined according to the deviation between the second simulation results and the true results.
[0061] Specifically, the accuracy of the second simulation results may be the reciprocal of the deviation between the second simulation results and the true results. The larger the deviation, the smaller the accuracy of the second simulation results. The deviation may be determined by means such as absolute deviation and relative deviation.
[0062] In one embodiment, the number of the first simulation results is multiple, including multiple first simulation results output by the crop simulation models corresponding to each initial genetic parameter group after multiple groups of crop planting data are respectively input into the crop simulation models corresponding to each initial genetic parameter group. The accuracy of the first simulation results is determined according to the deviation between the multiple first simulation results and the true results corresponding to each first simulation result.
[0063] It can be understood that there may be multiple groups of crop planting data. After multiple groups of crop planting data are respectively input into the crop simulation models corresponding to each initial genetic parameter group, the crop simulation models corresponding to each initial genetic parameter group may output corresponding first simulation results for each group of crop planting data, so that multiple first simulation results corresponding to each initial genetic parameter group can be obtained. The accuracy of the first simulation results may be determined according to the deviation between the multiple first simulation results and the true results corresponding to each first simulation result (i.e., the true results corresponding to each group of crop planting data). For example, it may be determined according to the root mean square error (RMSE) and / or the normalized root mean square error (NRMSE) between the first simulation results and the true results corresponding to each first simulation result. The normalized root mean square error helps to compare datasets or models of different scales.
[0064] In one embodiment, the number of the second simulation results is multiple, including multiple second simulation results output by the crop simulation models corresponding to each fine-tuned genetic parameter group after multiple groups of crop planting data are respectively input into the crop simulation models corresponding to each fine-tuned genetic parameter group. The accuracy of the second simulation results is determined according to the deviation between the multiple second simulation results and the true results corresponding to each second simulation result.
[0065] It can be understood that there can be multiple sets of crop planting data. After the multiple sets of crop planting data are respectively input into the crop simulation models corresponding to each fine-tuned genetic parameter group, the crop simulation models corresponding to each fine-tuned genetic parameter group can output corresponding second simulation results for each set of crop planting data, so that multiple second simulation results corresponding to each fine-tuned genetic parameter group can be obtained. The accuracy of the second simulation results can be determined according to the deviation between the multiple second simulation results and the true results corresponding to each second simulation result (i.e., the true results corresponding to each set of crop planting data). For example, it can be determined according to the root mean square error (RMSE) and / or normalized root mean square error (NRMSE) between the second simulation results and the true results corresponding to each second simulation result. The normalized root mean square error helps to compare datasets or models of different scales.
[0066] In one embodiment, when the number of the first simulation results corresponding to each initial genetic parameter group is multiple, the accuracy of the first simulation results can be determined according to the deviation between the first simulation results and the true results, where the deviation can be determined by means such as standard deviation, variance, mean square deviation, mean difference, Z-test, T-test, and regression analysis.
[0067] In one embodiment, when the number of the second simulation results corresponding to each fine-tuned genetic parameter group is multiple, the accuracy of the second simulation results can be determined according to the deviation between the second simulation results and the true results, where the deviation can be determined by means such as standard deviation, variance, mean square deviation, mean difference, Z-test, T-test, and regression analysis.
[0068] In one embodiment, according to the adjustment parameter value ranges of each genetic parameter, multiple fine-tuned genetic parameter groups are determined, including: according to the adjustment parameter value ranges of each genetic parameter, multiple fine-tuned genetic parameter groups are determined by means of exhaustive search.
[0069] Specifically, the processor can determine multiple fine-tuned genetic parameter groups by means of exhaustive search according to the adjustment parameter value ranges of each genetic parameter, which can prevent data omission, make the data more comprehensive and reliable, and further improve the accuracy of the parameter adjustment results.
[0070] In one embodiment, the output parameters of the crop simulation model can include growth period parameters and / or yield parameters.
[0071] In one embodiment, the genetic parameters can include but are not limited to: vegetative period, critical photoperiod, degree of developmental delay, time from the start of grain filling to physiological maturity, potential spikelet number coefficient, single grain weight, tillering coefficient, phylloid interval, temperature at which spikelet sterility is affected by high temperature, temperature that delays spike differentiation, and temperature that affects spikelet sterility.
[0072] Specifically, the genetic parameters may include the vegetative period P1, the critical photoperiod P2O, the degree of developmental delay P2R, the time from the start of grain filling to physiological maturity P5, the potential spikelet number coefficient G1, the single grain weight G2, the tillering coefficient G3, the phyllochron PHINT, the temperature THOT at which spikelet sterility is affected by high temperature, the temperature TCLDP that delays panicle differentiation, and the temperature TCLDF that affects spikelet sterility. Among them, the vegetative period P1 refers to the time period from emergence to when the rice plant is insensitive to photoperiod changes (in °C*d above a base temperature of 9°C, expressed as growing degree days [GDD]). This period is also known as the basic vegetative period of the plant. The suitable temperature range for this period is 150 - 800 °C*d. It can be calibrated by comparing with the observed panicle initiation and flowering dates. The critical photoperiod P2O refers to the critical photoperiod or the longest day length (in hours) at which development occurs at the maximum rate. At values above P2O, the development rate slows down, and thus it is delayed due to longer day lengths. The critical photoperiod range is generally 11 - 13 hours. The default value is 12 hours. When calibrating, it is generally not less than 11 hours. The degree of developmental delay P2R refers to the degree of stage - by - stage developmental delay of panicle differentiation caused by each additional hour of photoperiod above P2O (expressed as GDD in °C*d). The temperature range for this stage is 5 - 300 °C*d. Values for modern rice varieties will be in the lower range. The time from the start of grain filling to physiological maturity P5 refers to the period (GDD °C*d) from the start of grain filling (3 to 4 days after flowering) to physiological maturity (with a base temperature of 9°C). The temperature range for this stage is 150 - 850 °C*d. When calibrating, first ensure that P1, P2O, and P2R are correctly calibrated against the flowering data. Then calibrate P5 based on the observed maturity date. The potential spikelet number coefficient G1 can be estimated based on the number of spikelets per gram of main stem dry weight at flowering (less leaves and leaf sheaths plus spikelets). The range is 50 - 75 # / g, and the typical value is 55 # / g. The single grain weight G2 refers to the weight of a single grain (g) under ideal growth conditions, i.e., with unlimited light, water, nutrients, and freedom from pests and diseases. The range of G2 is generally 0.015 - 0.030 grams, and the default value is 0.025 grams. G2 has very low flexibility. The tillering coefficient G3 refers to the tillering coefficient relative to the IR64 variety under ideal conditions (scalar value). The range of G3 is generally 0.7 - 1.3. The coefficient for varieties with a higher tillering rate can be greater than 1.0. The phyllochron PHINT refers to the time interval (in degree - days) between the appearance of each leaf tip under non - stress conditions. The temperature range is generally 55 - 90 degrees Celsius, and the default value is 83 degrees Celsius. The temperature range of the temperature THOT at which spikelet sterility is affected by high temperature is generally 25 - 34 degrees Celsius, and the default temperature is 28 degrees Celsius. In specific operations, unless thermal environment data is available, it is recommended not to change THOT. For old varieties, THOT = 28 / G4.The temperature TCLDP that delays panicle differentiation refers to the temperature below which low temperature will further delay panicle differentiation (except for P1, P2O, and P2R). The temperature range of TCLDP is 12 - 18 degrees Celsius, and the default temperature is 15 degrees Celsius. In specific operations, it is recommended not to change TCLDP unless cold environment data is available. For converting old varieties, TCLDP = 15 * G5. The temperature TCLDF that affects spikelet sterility refers to the temperature (°C) at which low temperature affects spikelet sterility. The temperature range is 10 - 20 degrees Celsius. The default temperature is 15 degrees Celsius. In specific operations, it is recommended not to change TCLDF unless cold environment data is available. For converting old varieties, TCLDF = 15 * G5.
[0073] In one embodiment, the number of multiple initial genetic parameter groups is within a preset number range.
[0074] It can be understood that the preset number range is the number of initially set genetic parameter groups. For example, the number range can be 1000 - 5000. Setting the number of initial genetic parameter groups within a reasonable range can reduce the parameter tuning duration, improve the parameter tuning efficiency, and at the same time improve the accuracy of the parameter tuning results.
[0075] In one embodiment, crop planting data may include but is not limited to data such as basic data, fertilization data, irrigation data, soil data, observation data, and meteorological data. Among them, the basic data may include longitude and latitude, crop, variety, sowing time, transplanting time, row spacing, planting depth, transplanting temperature, and seed amount per mu. The fertilization data may include fertilization time, fertilization name (nitrogen, phosphorus, and potassium content), fertilization method, fertilization depth, and fertilization amount. The irrigation data may include irrigation time, irrigation method, and water layer depth. The soil data may include sampling time, soil texture, sand content, PH measurement method and value, phosphorus measurement method and value, potassium measurement method and value, total nitrogen, total phosphorus, available potassium, and organic carbon. The observation data may include dry rice yield per mu, initial stage of young panicle differentiation, initial stage of flowering, and initial stage of maturity. The meteorological data may include date, solar radiation, daily minimum temperature, daily maximum temperature, and daily rainfall.
[0076] In a specific embodiment, the parameter tuning process in the parameter tuning method for a crop simulation model may include two main steps: rough tuning and fine tuning. The rough tuning step may include the following steps:
[0077] (1) Initialize the value range of each genetic parameter in the initial genetic parameter group (i.e., the corresponding initial parameter value range) to generate an initialized parameter range configuration file (corresponding to the initial model parameters).
[0078] (2) Select sample data from the rice planting database: Based on the actual data indicators of dry rice yield per mu, initial stage of young panicle differentiation, initial stage of flowering, and initial stage of maturity in the sample data, prepare for the comparative analysis of the next calculation results.
[0079] (3) Select the GLUESelect method to adjust parameters:
[0080] a) Set the calibration parameters for the results (corresponding to the real data), divide the genetic parameters into growth stage-related parameters and yield-related parameters, and support selecting any type of parameter or all parameters;
[0081] b) Set the number of runs number, usually set to 1000 - 5000 times (corresponding to the number of initial genetic parameter groups);
[0082] c) Set the accuracy percentage of result analysis data cleaning (corresponding to the simulation result accuracy).
[0083] (4) Calculate through the integrated GLUE module: According to the initialized parameter range and the set calibration parameters, use the glue to find the optimal solution algorithm to obtain the results of each operation, repeat the operation the number of times set in the previous step, and output the result file data for running number times.
[0084] (5) File data access and parsing: Extract the GLUE model operation result file data to the big data source layer ODS using the etl tool.
[0085] (6) Data model: Support three-dimensional analysis of the combination of growth stage parameters and yield parameters, use Spark to analyze the RMSE and NRMSE of the growth stage yield of the simulation results and real data, and obtain multiple groups of parameters with relatively high accuracy.
[0086] The fine-tuning steps may include the following steps:
[0087] (1) Parameter group parameter range: Based on the rough-tuning results of the previous step, select N parameter groups with relatively high accuracy (N is greater than or equal to 1), and obtain the maximum and minimum values of all genetic parameters in the N parameter groups. Through this step, the value range of each parameter is further narrowed compared to the initialized range.
[0088] (2) Parameter group parameter step size: Based on the maximum and minimum values of all genetic parameters obtained in the previous step, further set the upper and lower floating ranges of the parameters, that is, the floating step size of the parameters.
[0089] (3) Adjust parameters by the exhaustive method. According to the maximum, minimum, and step size of all genetic parameters set, obtain all parameter combinations, run each parameter combination through the dssat model once, and save the result file for each time.
[0090] (4) File data access and parsing: Extract the GLUE model operation result file data to the big data source layer ODS using the etl tool.
[0091] (5) Data model: Support three-dimensional analysis of growth period parameters and yield parameter combinations. Use Spark to analyze the RMSE and NRMSE of the growth period yield of the simulation results and real data, and obtain multiple groups of parameters with higher accuracy.
[0092] (6) If the parameter groups obtained after one rough adjustment and fine adjustment are average, it may be related to insufficient experimental sample data. At the same time, this solution also supports repeated fine adjustment until parameter groups that meet the requirements are obtained.
[0093] The above technical solution integrates the DSSAT module and combines a parameter tuning method that combines GLUESelect rough adjustment and exhaustive fine adjustment, improving the accuracy and flexibility of debugging and reducing the cost of parameter tuning for researchers. By combining the operation of a large number of experiments and big data analysis, it saves time costs. Using big data analysis capabilities, it is fast, accurate, and convenient, and can achieve rough and fine adjustment of crop model genetic parameters, and combines big data analysis capabilities to assign values using computer and big data capabilities.
[0094] An embodiment of the present invention provides a processor configured to execute the parameter tuning method for a crop simulation model according to the above embodiments.
[0095] An embodiment of the present invention provides a parameter tuning device for a crop simulation model, including: a processor according to the above embodiments.
[0096] An embodiment of the present invention provides a machine-readable storage medium, on which instructions are stored, and the instructions are used to cause the machine to execute the parameter tuning method for a crop simulation model according to the above embodiments.
[0097] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0098] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processors of general-purpose computers, special-purpose computers, embedded processors, or other programmable data processing devices to generate a machine, such that the instructions executed by the processors of the computer or other programmable data processing devices produce means for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 or means for implementing the functions specified in multiple blocks.
[0099] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including instruction means that implement the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 or means for implementing the functions specified in multiple blocks.
[0100] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 or means for implementing the functions specified in multiple blocks.
[0101] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and a memory.
[0102] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM), and / or non-volatile memory, such as read-only memory (ROM) or flash memory (flash RAM). The memory is an example of computer-readable media.
[0103] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.
[0104] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0105] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.
Claims
1. A method for adjusting parameters of a crop simulation model, characterized in that, The parameter adjustment method includes: Obtain the input parameters and initial model parameters of the crop simulation model. Among them, the input parameters include crop planting data, and the initial model parameters include multiple initial genetic parameter groups. The values of each genetic parameter in each initial genetic parameter group do not exceed the initial parameter value range of each genetic parameter. According to the crop planting data and the multiple initial genetic parameter groups, determine the first simulation result accuracy of the crop simulation model under each initial genetic parameter group. Screen the multiple initial genetic parameter groups according to the first simulation result accuracy to obtain multiple roughly adjusted genetic parameter groups. Among them, the roughly adjusted genetic parameter group is the initial genetic parameter group in the multiple initial genetic parameter groups whose first simulation result accuracy reaches the first preset accuracy. Determine the maximum and minimum values of each genetic parameter in the multiple roughly adjusted genetic parameter groups to obtain the adjustment parameter value range of each genetic parameter. According to the adjustment parameter value range of each genetic parameter, determine multiple finely adjusted genetic parameter groups. According to the crop planting data and the multiple finely adjusted genetic parameter groups, determine the second simulation result accuracy of the crop simulation model under each finely adjusted genetic parameter group. Determine the target genetic parameter group in the multiple finely adjusted genetic parameter groups according to the second simulation result accuracy to obtain the final model parameters of the crop simulation model. Among them, the target genetic parameter group is the finely adjusted genetic parameter group in the multiple finely adjusted genetic parameter groups whose second simulation result accuracy reaches the second preset accuracy, and the second preset accuracy is greater than the first preset accuracy.
2. The parameter adjustment method according to claim 1, wherein The step of determining the first simulation result accuracy of the crop simulation model under each initial genetic parameter group according to the crop planting data and the multiple initial genetic parameter groups includes: Input the crop planting data into the crop simulation model corresponding to each initial genetic parameter group to obtain the first simulation result output by the crop simulation model corresponding to each initial genetic parameter group. According to the first simulation result and the true result corresponding to the pre-obtained crop planting data, determine the first simulation result accuracy of the crop simulation model under each initial genetic parameter group. The step of determining the second simulation result accuracy of the crop simulation model under each finely adjusted genetic parameter group according to the crop planting data and the multiple finely adjusted genetic parameter groups includes: Input the crop planting data into the crop simulation model corresponding to each finely adjusted genetic parameter group to obtain the second simulation result output by the crop simulation model corresponding to each finely adjusted genetic parameter group. According to the second simulation result and the true result corresponding to the pre-obtained crop planting data, determine the second simulation result accuracy of the crop simulation model under each finely adjusted genetic parameter group.
3. The parameter adjustment method according to claim 2, wherein The first simulation result accuracy is determined according to the deviation between the first simulation result and the true result; the second simulation result accuracy is determined according to the deviation between the second simulation result and the true result.
4. The parameter adjustment method according to claim 2, wherein The number of the first simulation results is multiple, including multiple first simulation results output by the crop simulation models corresponding to each of the initial genetic parameter groups after multiple groups of the crop planting data are respectively input into the crop simulation models corresponding to each of the initial genetic parameter groups, and the accuracy of the first simulation results is determined according to the deviation between the multiple first simulation results and the true results corresponding to each of the first simulation results; The number of the second simulation results is multiple, including multiple second simulation results output by the crop simulation models corresponding to each of the fine-tuned genetic parameter groups after multiple groups of the crop planting data are respectively input into the crop simulation models corresponding to each of the fine-tuned genetic parameter groups, and the accuracy of the second simulation results is determined according to the deviation between the multiple second simulation results and the true results corresponding to each of the second simulation results.
5. The parameter adjustment method according to claim 1, characterized in that, Determining multiple fine-tuned genetic parameter groups according to the adjustment parameter value ranges of the respective genetic parameters, includes: Determining multiple fine-tuned genetic parameter groups by using the exhaustive method according to the adjustment parameter value ranges of the respective genetic parameters.
6. The parameter adjustment method according to claim 1, wherein The output parameters of the crop simulation model include growth period parameters and / or yield parameters.
7. The parameter adjustment method according to claim 1, wherein The number of the multiple initial genetic parameter groups is within a preset number range.
8. A processor, characterized in that, Configured to execute the parameter adjustment method for a crop simulation model according to any one of claims 1 to 7.
9. A parameter adjustment device for a crop simulation model, characterized in that, Includes: The processor according to claim 8.
10. A machine-readable storage medium, characterized in that, Instructions are stored on the machine-readable storage medium, and the instructions are used to cause the machine to execute the parameter adjustment method for a crop simulation model according to any one of claims 1 to 7.