Evaluation method for winter wheat overwintering freeze injury death rate

Abstract

The invention relates to the field of wheat overwintering freezing injury death rate evaluation, in particular to a winter wheat overwintering freezing injury death rate evaluation method which comprises the steps that a first-stage cold-resistant action factor and a second-stage cold-resistant action factor are obtained respectively; if the wheat tillering joint temperature exceeds a preset threshold value, calculating a cold-resistant exercise relieving rate; obtaining a critical anti-freezing temperature of the wheat variety one day before evaluation; constructing a low-temperature freeze injury prediction model; determining the prediction accuracy, the root mean square error, the real deviation value and the significant difference value of the low-temperature freeze injury prediction model; judging whether the simulation accuracy of the cold-resistant exercise relieving effect meets the requirement or not, and determining different cold-resistant exercise response modes according to the simulation accuracy result. According to the method, the dynamic change of the anti-freezing capacity of different varieties of wheat in the overwintering process is detected through the cold-resistant exercise process and the cold-resistant exercise relieving process, and the simulation accuracy of the low-temperature freezing injury prediction model on different types of freezing injuries is enhanced.

Description

Technical Field

[0001] The invention relates to the field of evaluation of mortality rate of winter wheat due to freezing damage, and in particular to an evaluation method for mortality rate of winter wheat due to freezing damage. Background Art

[0002] Wheat is the third largest food crop in the world after corn and rice, with an annual global output of more than 6 billion tons. 21% of the world's food supply depends on wheat production. At present, the research focus on wheat frost damage is mainly on the mechanism of the impact of frost damage on wheat growth and development, the morphology of plants after the disaster, the maximum frost resistance temperature index in midwinter, and the corresponding risk assessment. With the combination and cross-penetration of information science and agricultural science, research on agricultural information technology at home and abroad has made significant progress. However, crop simulation technology is not yet perfect, the description of some plant physiological and ecological processes is still lacking in scientificity, and there is a lack of mechanistic research on the simulation of crop disasters. The existing crop models that have achieved a good application level are at the third level of crop production level, especially the simulation module of the impact of disaster factors is still insufficient or lacking. Among them, there is little research on the wheat frost damage impact model module, which seriously affects the applicability of crop models.

[0003] Chinese patent publication number CN108414695A discloses a wheat cold resistance simulation identification device, including a flower pot, a bracket, a mesh, a rope, a wheat plant, a refrigerator, a refrigeration switch, a soil layer, a solar panel, a battery, a temperature sensor, a sensor bracket and a controller; a soil layer is arranged in the flower pot, and the wheat plant is planted in the center of the soil layer; the bottom of the bracket is arranged around the mouth of the flower pot, and the other end of the bracket extends to a height sufficient to accommodate the growth of the wheat plant; the bottom of the sensor bracket is arranged next to the wheat plant in the center of the soil layer, so that the top of the sensor bracket is located in the middle of the height of the bracket, and the top of the sensor bracket is provided with a temperature sensor. The device of the present invention has a simple structure, intuitive identification object comparison, and obvious effect; the solar panel is used to supplement the power in the battery, eliminating the step of repeated charging, and is clean and environmentally friendly; the mesh can avoid the influence of other natural factors, and the identification result is more accurate and reliable. It can be seen that the prior art has the problem of being unable to detect the dynamic changes in the antifreeze ability of different varieties of wheat during the wintering process, thereby causing the low temperature frost damage prediction model to simulate different types of frost damage inaccurately. Summary of the invention

[0004] To this end, the present invention provides a method for evaluating the mortality rate of winter wheat overwintering frost damage, so as to overcome the problem that the prior art is unable to detect the dynamic changes in the frost resistance of different varieties of wheat during the wintering, thereby leading to inaccurate simulation of different types of frost damage by low temperature frost damage prediction models.

[0005] To achieve the above object, the present invention provides a method for evaluating winter wheat frost injury mortality during winter, comprising:

[0006] The first stage and the second stage of cold-resistance training of wheat are simulated respectively to obtain the first stage cold-resistance factor and the second stage cold-resistance factor;

[0007] If the temperature of the wheat tillering node exceeds a preset threshold, the critical antifreeze temperature of the wheat that has not been hardened against cold is detected to determine the cold hardening release rate in combination with the critical antifreeze temperature of the wheat that has been hardened against cold;

[0008] Obtaining the critical antifreeze temperature of the wheat one day before the evaluation;

[0009] Constructing a low temperature frost damage prediction model according to the first-stage cold resistance factor, the second-stage cold resistance factor, the cold resistance training release rate, and the critical cold resistance temperature on the day before the evaluation;

[0010] Calculate the observed value, simulated value, average value of the observed value, average value of the simulated value and the number of samples of the low temperature frost damage prediction model to determine the prediction accuracy, root mean square error, true deviation and significant difference of the low temperature frost damage prediction model;

[0011] Determining the simulation accuracy of the cold hardship release effect according to the true deviation to determine whether to adjust the number of segments of the vernalization temperature or to adjust the threshold of the tillering node temperature of the cold hardship release effect according to the average difference between the critical cold hardiness temperatures of the root and stem of the wheat and the leaves after thawing;

[0012] If the simulation accuracy still does not meet the requirements after the cold-resistant training coping method is executed, the ratio of data training to data replacement is adjusted according to the EF value of the low-temperature frost damage prediction model.

[0013] Furthermore, the calculation formula of the low temperature damage prediction model is:

[0014] LT i =LT i-1 +f hardningi +f dehardningi

[0015] Among them, LT i is the critical antifreeze temperature threshold on the i-th day during the wintering process, LT i-1 It is the critical antifreeze temperature of the variety the day before. hardningi is the cold resistance training rate on the i-th day, f dehardningi is the rate of cold resistance training release on day i;

[0016] The lowest air temperature of the day is T min <LT i, it is determined that wheat has frost damage, when T min >LT i No frost damage occurs.

[0017] Furthermore, the calculation formula of the cold resistance factor in the first stage is:

[0018] f hardningi1 =f la ×ΔPot h1 ×min(f ti1 ,f shi )

[0019] Among them, f hardningi1 is the change in wheat temperature in the first stage of cold hardiness training, f ti1 is the temperature effect factor on the ith day in the first stage of cold-resistance training, f shi is the light effect factor on the i-th day, f la is the pre-winter leaf age factor, with a value between 0 and 1, ΔPot h1 is the potential increase in daily wheat temperature after cold hardiness.

[0020] Furthermore, the calculation formula of the cold resistance factor in the second stage is:

[0021]

[0022] Among them, f hardningi2 is the change of wheat temperature in the second stage of cold hardening, HF2 is the cold resistance effect factor in the first stage, R c1 R is the critical temperature of wheat frost resistance after the first stage of cold resistance training. c2 LT is the critical temperature of wheat frost resistance after the second stage of cold resistance training. i is the critical antifreeze temperature threshold on the i-th day during the wintering process, L max It is the critical antifreeze temperature of wheat that has been hardened to cold.

[0023] Furthermore, the calculation formula for the prediction accuracy of the low temperature frost damage prediction model is:

[0024]

[0025] Among them, EF>0 means that the prediction accuracy of the low temperature frost damage prediction model meets the requirements;

[0026] The root mean square error of the low temperature damage prediction model is the absolute error between the simulated value and the measured value, and the calculation formula is:

[0027]

[0028] Among them, RMSE ais the root mean square error of the low temperature damage prediction model;

[0029] The true deviation of the low temperature damage prediction model is the sum of the deviations between the simulated value and the measured value, and the calculation formula is:

[0030]

[0031] Among them, R 2 is the true deviation of the low temperature damage prediction model;

[0032] The significant difference of the low temperature damage prediction model is the difference between the simulated value and the measured value, and the calculation formula is:

[0033] ABS=S i -O i ;

[0034] Among them, ABS is the significant difference of low temperature damage prediction model;

[0035] Among them, O i is the observed value, S i is the analog value, is the mean of the observed values, is the average value of the simulation, and n is the number of samples.

[0036] Furthermore, whether the simulation accuracy of the cold-resistant training release effect meets the requirements is determined based on the actual deviation of the low-temperature frost damage prediction model. When the actual deviation of the low-temperature frost damage prediction model is greater than the preset second deviation, it is determined that the simulation accuracy of the cold-resistant training release effect does not meet the requirements, and the number of segments of the vernalization temperature is increased.

[0037] Furthermore, when the actual deviation of the low-temperature frost damage prediction model is greater than the preset first deviation and less than or equal to the preset second deviation, it is determined that the simulation accuracy of the cold-hardening release effect does not meet the requirements, and a secondary judgment is made on the simulation accuracy of the cold-hardening release effect based on the average difference in the critical cold-hardiness temperature of the rhizomes and leaves of the wheat after thawing. When the average difference in the critical cold-hardiness temperature of the rhizomes and leaves of the wheat after thawing is greater than the preset temperature difference, the tillering node threshold of the cold-hardening release effect is adjusted.

[0038] Furthermore, if the actual deviation of the low-temperature frost damage prediction model is still greater than the preset second deviation after adjustment, the simulation accuracy of the cold-resistant training release effect is determined according to the EF value of the low-temperature frost damage prediction model. If the EF value of the low-temperature frost damage prediction model is less than or equal to the EF value of the low-temperature frost damage model, it is determined that the simulation accuracy of the cold-resistant training release effect does not meet the requirements, and the ratio of data training to data replacement is adjusted.

[0039] Furthermore, the average difference in critical cold-resistance temperatures of the rhizomes and leaves of the thawed wheat is the difference between the critical temperatures at which the rhizomes and leaves of the thawed wheat can withstand low temperatures without suffering from frost damage; the wheat tillering node temperature refers to the extreme temperature condition that affects the growth and development of the wheat tillering node.

[0040] Furthermore, the ratio of data training to data replacement is the ratio of the amount of data used for training optimization problems in the data with problems when the low temperature damage prediction model is running to the amount of data used for search and replacement in the database;

[0041] Among them, the sum of the data volume for the training optimization problem and the data volume for the search and replacement in the database is the data volume of data with problems when the low-temperature frost damage prediction model is running.

[0042] Compared with the prior art, the beneficial effects of the present invention are that, by simulating two stages of wheat cold hardening respectively, the present invention can consider the formation and change of wheat cold resistance at different stages in a more detailed and accurate manner, thereby improving the understanding of its cold resistance characteristics; by calculating the cold hardening release rate based on the tillering node temperature exceeding a preset threshold, the accuracy of capturing key nodes that affect the change of wheat cold resistance can be improved; by combining cold resistance factors, cold hardening release rates and critical antifreeze temperatures at different stages to construct a low temperature frost damage prediction model, a number of key influencing factors are comprehensively considered, thereby improving the comprehensiveness of the prediction model; the present invention helps farmers take effective preventive measures in advance through frost damage prediction, reduces frost damage losses, ensures wheat yield and quality, provides a scientific decision-making basis for agricultural production, and provides effective evaluation means and data support for screening and breeding of varieties with stronger frost resistance.

[0043] Furthermore, the present invention overcomes the problem of inaccurate wheat development process algorithm caused by inaccurate data sampling during wheat development process simulation, which in turn leads to inaccurate vernalization temperature, by setting the real deviation amount of the low temperature frost damage prediction model. When it is determined that the simulation does not meet the requirements, increasing the number of segments of the vernalization temperature helps to more carefully consider the impact of temperature changes on wheat frost mortality, thereby improving the accuracy and reliability of the model; by determining the accuracy of the frost mortality simulation based on the real deviation amount, it is possible to promptly discover deviations in the simulation results, improve and optimize the model, and improve the adaptability of the prediction model under different environments and planting conditions.

[0044] Furthermore, the present invention overcomes the inaccuracy of the cold-resistance training release effect caused by part of the water being absorbed by the soil after thawing due to the sudden change of water, which leads to a slowdown in growth or failure to meet growth conditions, and the thawing of the surface soil. The roots and stems of some wheat are exposed and interact with the environment, thereby triggering different levels of cold-resistance training in some areas, and performing a secondary judgment on the cold-resistance training release effect within the range of a preset first deviation and a preset second deviation. The accuracy of the model can be evaluated more carefully, which helps to improve the accuracy of the judgment model, thereby further improving the simulation effect of the model.

[0045] Furthermore, the present invention sets the EF value of the low-temperature frost damage prediction model. When the critical cold-resistance temperature difference of wheat is still too large, further judgment is made with the help of the EF value, which helps to more accurately find out the problems with the model. By adjusting the ratio of data training to data replacement, the model's ability to learn and adapt to new data can be improved, thereby improving the model's generalization ability and prediction accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 This is an overall step diagram of a method for evaluating winter wheat overwintering freeze injury mortality according to an embodiment of the present invention;

[0047] Figure 2 A line graph showing the effect of pre-winter leaf age on the rate of cold-resistance training of wheat according to an embodiment of the present invention;

[0048] Figure 3 A line graph showing the actual deviation of the critical antifreeze temperature of wheat during wintering according to an embodiment of the present invention;

[0049] Figure 4 The present invention is a flowchart of a method for evaluating the mortality rate of winter wheat caused by freezing damage during the winter. DETAILED DESCRIPTION

[0050] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0051] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0052] It will be understood by those skilled in the art that, unless otherwise stated, the singular forms "a", "an" and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in this specification refers to the presence of features, integers, steps, operations, elements / components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements / components. It should be understood that when we refer to a module as being "connected" or "coupled" to another module, it may be directly connected or coupled to the other module, or there may be an intermediate unit. In addition, "connected" or "coupled" used herein may include wireless connection or wireless coupling.

[0053] See also Figure 1 , Figure 2 , Figure 3 as well as Figure 4 As shown, they are respectively an overall step diagram of the method for evaluating the mortality rate of winter wheat freezing injury in winter according to an embodiment of the present invention, a line graph of the influence of pre-winter leaf age on the rate of wheat cold resistance training, a line graph of the actual deviation of the critical freezing temperature during the wintering of wheat, and a flow chart. The method for evaluating the mortality rate of winter wheat freezing injury in winter according to the present invention comprises:

[0054] The first stage and the second stage of cold-resistance training of wheat are simulated respectively to obtain the first stage cold-resistance factor and the second stage cold-resistance factor;

[0055] If the temperature of the wheat tillering node exceeds a preset threshold, the critical antifreeze temperature of the wheat that has not been hardened against cold is detected to determine the cold hardening release rate in combination with the critical antifreeze temperature of the wheat that has been hardened against cold;

[0056] Obtaining the critical antifreeze temperature of the wheat one day before the evaluation;

[0057] Constructing a low temperature frost damage prediction model according to the first-stage cold resistance factor, the second-stage cold resistance factor, the cold resistance training release rate, and the critical cold resistance temperature on the day before the evaluation;

[0058] Calculate the observed value, simulated value, average value of the observed value, average value of the simulated value and the number of samples of the low temperature frost damage prediction model to determine the prediction accuracy, root mean square error, true deviation and significant difference of the low temperature frost damage prediction model;

[0059] Determining the simulation accuracy of the cold hardship release effect according to the true deviation to determine whether to adjust the number of segments of the vernalization temperature or to adjust the threshold of the tillering node temperature of the cold hardship release effect according to the average difference between the critical cold hardiness temperatures of the root and stem of the wheat and the leaves after thawing;

[0060] If the simulation accuracy still does not meet the requirements after the cold-resistant training coping method is executed, the ratio of data training to data replacement is adjusted according to the EF value of the low-temperature frost damage prediction model.

[0061] In practice, the present invention can consider the formation and change of wheat's cold resistance at different stages in a more detailed and accurate manner by simulating two stages of wheat cold training respectively, thereby improving the understanding of its cold resistance characteristics; by setting the cold training release rate to be calculated based on the tillering node temperature exceeding a preset threshold, the accuracy of capturing key nodes that affect the change of wheat's cold resistance can be improved; by combining the cold resistance factors, cold training release rate and critical antifreeze temperature at different stages to construct a low temperature frost damage prediction model, a number of key influencing factors are comprehensively considered, thereby improving the comprehensiveness of the prediction model; the present invention helps farmers take effective preventive measures in advance through frost damage prediction, reduces frost damage losses, ensures wheat yield and quality, provides a scientific decision-making basis for agricultural production, and provides effective evaluation means and data support for screening and breeding of varieties with stronger frost resistance.

[0062] Specifically, the calculation formula of the low temperature frost damage prediction model is:

[0063] LT i =LT i-1 +f hardningi +f dehardningi

[0064] Among them, LT i is the critical antifreeze temperature threshold on the i-th day during the wintering process, LT i-1 It is the critical antifreeze temperature of the variety the day before. hardningi is the cold resistance training rate on the i-th day, f dehardningi is the rate of cold resistance training release on day i;

[0065] The lowest air temperature of the day is T min <LT i , it is determined that wheat has frost damage, when T min >LT i No frost damage occurs.

[0066] Specifically, f hardningi Affected by variety, temperature and light, it promotes the improvement of wheat's frost resistance; dehardningi It is mainly affected by variety type and temperature, and has an inhibitory effect on wheat's frost resistance.

[0067] Specifically, the calculation formula of the first stage cold resistance factor is:

[0068] f hardningi1 =f la ×ΔPoth1 ×min(f ti1 ,f shi )

[0069] Among them, f hardningi1 is the change in wheat temperature in the first stage of cold hardiness training, f ti1 is the temperature effect factor on the ith day in the first stage of cold-resistance training, f shi is the light effect factor on the i-th day, f la is the pre-winter leaf age factor, with a value between 0 and 1, ΔPot h1 is the potential increase in daily wheat temperature after cold hardiness.

[0070] Specifically, the temperature effect factor f on the i-th day in the first stage of the cold resistance training is ti1 The calculation formula is:

[0071]

[0072] Among them, T a The temperature during the first stage of cold-resistance training.

[0073] Specifically, this study assumes that when the sunshine hours are less than 5 hours, it is considered that the sunshine is insufficient, which in turn affects the cold resistance of wheat; the light effect factor f on the i-th day is shi The calculation formula is:

[0074]

[0075] Among them, ssd is the number of sunshine hours on that day.

[0076] Specifically, the pre-winter leaf age factor f la The calculation formula is:

[0077]

[0078] Among them, L a It is the leaf age before winter.

[0079] Specifically, the calculation formula of the second stage cold resistance factor is:

[0080]

[0081] Among them, f hardningi2 is the change of wheat temperature in the second stage of cold hardening, HF2 is the cold resistance effect factor in the first stage, R c1 R is the critical temperature of wheat frost resistance after the first stage of cold resistance training. c2 LT is the critical temperature of wheat frost resistance after the second stage of cold resistance training. iis the critical antifreeze temperature threshold on the i-th day during the wintering process, L max It is the critical antifreeze temperature of wheat that has been hardened to cold.

[0082] Specifically, after the first stage of cold-resistance training, the critical temperature of wheat freezing resistance R c1 The calculation formula is:

[0083]

[0084] Among them, N1 is the number of days in the first stage of cold resistance training.

[0085] Specifically, the change in wheat temperature during the second stage of cold hardiness training is f hardning2 The calculation formula is:

[0086] f hardning2 =[(L max -L c ) / N2]×f ti2

[0087] Among them, f ti2 Temperature effect factor of the second stage of cold resistance training, L c L is the maximum freezing temperature that a specific type of wheat can reach after the first stage of cold hardiness training. max is the critical freezing temperature of winter-hardened wheat, and N2 is the number of days in the second stage of winter-hardiness training;

[0088] When the temperature during the first stage of cold-resistance training is T a When the value range is [-5℃, 0℃), f ti2 The value is 1, when the temperature T a When the value range of is less than or greater than or equal to [-5℃, 0℃), f ti2 The value of is 0.

[0089] Specifically, if the temperature of the wheat tillering node exceeds a preset threshold, the cold hardening release rate is calculated according to the critical cold hardening temperature of the wheat that has been hardened;

[0090] The preset threshold value may range from [0°C to 10°C]. In a specific embodiment, the temperature of the wheat tillering node is 12°C, which exceeds the preset threshold value. The cold resistance training release rate f dehardning The calculation formula is:

[0091]

[0092] Among them, L max is the critical freezing temperature of wheat after cold hardening, L min is the critical freezing temperature of wheat that has not been hardened to cold, Ta The temperature during the first stage of cold-resistance training.

[0093] Specifically, the calculation formula for the prediction accuracy of the low temperature frost damage prediction model is:

[0094]

[0095] Among them, EF>0 means that the prediction accuracy of the low temperature frost damage prediction model meets the requirements;

[0096] The root mean square error of the low temperature damage prediction model is the absolute error between the simulated value and the measured value, and the calculation formula is:

[0097]

[0098] Among them, RMSE a is the root mean square error of the low temperature damage prediction model;

[0099] The true deviation of the low temperature damage prediction model is the sum of the deviations between the simulated value and the measured value, and the calculation formula is:

[0100]

[0101] Among them, R 2 is the true deviation of the low temperature damage prediction model;

[0102] The significant difference of the low temperature damage prediction model is the difference between the simulated value and the measured value, and the calculation formula is:

[0103] ABS=S i -O i ;

[0104] Among them, ABS is the significant difference of low temperature damage prediction model;

[0105] Among them, O i is the observed value, S i is the analog value, is the mean of the observed values, is the average value of the simulation, and n is the number of samples.

[0106] Specifically, whether the simulation accuracy of the cold-hardening release effect meets the requirements is determined according to the real deviation of the low-temperature frost damage prediction model. When the real deviation of the low-temperature frost damage prediction model is greater than the preset second deviation, it is determined that the simulation accuracy of the cold-hardening release effect does not meet the requirements, and the number of segments of the vernalization temperature is increased.

[0107] Among them, the vernalization temperature is divided into segments according to the temperature range required for winter wheat to complete the vernalization process. Generally, the vernalization temperature segmentation intervals are [0℃, 3℃], [3℃, 6℃] and [6℃, 15℃].

[0108] In implementation, the present invention overcomes the problem of inaccurate wheat development process algorithm caused by inaccurate data sampling during wheat development process simulation, which in turn leads to inaccurate vernalization temperature, by setting the real deviation of the low temperature frost damage prediction model. When it is determined that the simulation does not meet the requirements, increasing the number of segments of the vernalization temperature helps to more carefully consider the impact of temperature changes on wheat frost damage mortality, thereby improving the accuracy and reliability of the model; by determining the accuracy of the frost damage mortality simulation based on the real deviation, it is possible to promptly discover deviations in the simulation results, improve and optimize the model, and improve the adaptability of the prediction model under different environments and planting conditions.

[0109] Optionally, the preset second deviation value may be in the range of [5%, 10%];

[0110] Preferably, the preferred embodiment of the preset second deviation amount is 7%;

[0111] In a possible embodiment, the true deviation of the low temperature frost damage prediction model is 13%, which is greater than the preset second deviation. The simulation accuracy of determining the cold-resistant training release effect does not meet the requirements. For every 2% by which the true deviation of the low temperature frost damage prediction model is greater than the preset second deviation, the number of segments of the vernalization temperature increases by 1. At this time, the number of segments of the vernalization temperature is 3. The increased number of segments of the vernalization temperature is: 3+(13-7) / 2=6.

[0112] Specifically, when the actual deviation of the low-temperature frost damage prediction model is greater than the preset first deviation and less than or equal to the preset second deviation, it is determined that the simulation accuracy of the cold-hardening release effect does not meet the requirements, and a secondary judgment is made on the simulation accuracy of the cold-hardening release effect based on the average difference in the critical cold-hardiness temperature of the rhizomes and leaves of the wheat after thawing. When the average difference in the critical cold-hardiness temperature of the rhizomes and leaves of the wheat after thawing is greater than the preset temperature difference, the tillering node threshold of the cold-hardening release effect is adjusted.

[0113] In implementation, the present invention overcomes the inaccuracy of the cold-resistance training release effect caused by part of the water being absorbed by the soil after thawing due to the sudden change of water, which leads to a slowdown in growth or failure to meet growth conditions. The surface soil thaws, and some wheat roots and stems are exposed, interacting with the environment, thereby triggering different levels of cold-resistance training in some areas. A secondary judgment is made on the cold-resistance training release effect within the range of a preset first deviation and a preset second deviation, and the accuracy of the model can be evaluated more carefully, which helps to improve the accuracy of the judgment model, thereby further improving the simulation effect of the model.

[0114] Optionally, the preset first deviation value may be in a range of [2%, 5%], and the preset temperature difference value may be in a range of [3°C, 5°C];

[0115] Preferably, the preferred embodiment of the preset first deviation amount is 3%, and the preferred embodiment of the preset temperature difference amount is 4°C;

[0116] In a possible embodiment, the actual deviation of the low temperature frost damage prediction model is 6%, which is greater than the preset first deviation and less than the preset second deviation, and it is determined that the simulation accuracy of the cold-hardness training release effect does not meet the requirements; the average difference in the critical cold-hardness temperature of the rhizomes and leaves of the wheat after thawing is 6°C, which is greater than the preset temperature difference. For every 1°C increase in the average difference in the critical cold-hardness temperature of the rhizomes and leaves of the wheat after thawing that is greater than the preset first deviation, the maximum value of the tillering node threshold of the cold-hardness training release effect is reduced by 1°C. At this time, the tillering node threshold is [0°C, 10°C], and the maximum value of the reduced tillering node threshold is 10-(6-4)=8, and the reduced tillering node threshold is [0°C, 8°C].

[0117] Specifically, if the actual deviation of the low-temperature frost damage prediction model is still greater than the preset second deviation after adjustment, then the simulation accuracy of the cold-resistant training release effect is determined to meet the requirements according to the EF value of the low-temperature frost damage prediction model; if the EF value of the low-temperature frost damage prediction model is less than or equal to the EF value of the preset low-temperature frost damage model, then it is determined that the simulation accuracy of the cold-resistant training release effect does not meet the requirements, and the ratio of data training to data replacement is adjusted;

[0118] In a specific embodiment, the EF value of the preset low-temperature frost damage model is 0, the EF value of the low-temperature frost damage model is -1, and the simulation accuracy of the cold-resistant training release effect is determined to be unsatisfactory. When the EF value of the low-temperature frost damage model is less than the EF value of the low-temperature frost damage model by 0.5, the proportion of data training is increased by 1% compared with the current data training proportion. The current data training proportion is 34%, and the increased data training proportion is 34% + |-1 / 0.5| = 36%.

[0119] In implementation, the present invention sets the EF value of the low-temperature frost damage prediction model. When the critical cold-resistance temperature difference of wheat is still too large, further judgment is made with the help of the EF value, which helps to more accurately find out the problems of the model. By adjusting the ratio of data training to data replacement, the model's ability to learn and adapt to new data can be improved, thereby improving the model's generalization ability and prediction accuracy.

[0120] Specifically, the average difference in critical cold resistance temperatures of the rhizomes and leaves of the thawed wheat is the average value of the difference between the critical temperatures at which the rhizomes and leaves of the thawed wheat can withstand low temperatures without suffering from frost damage; the wheat tillering node temperature refers to the extreme temperature conditions that affect the growth and development of the wheat tillering nodes.

[0121] Specifically, the ratio of data training to data replacement is the ratio of the amount of data used for training optimization problems in the data with problems when the low temperature damage prediction model is running to the amount of data used for search and replacement in the database;

[0122] The sum of the amount of data for the training optimization problem and the amount of data for search and replacement in the database is the amount of data that has problems when the low temperature damage prediction model is running;

[0123] Among them, the data with problems when the low-temperature frost damage prediction model is running are the data used by the prediction model in the prediction process when the low-temperature frost damage prediction model has prediction errors. For example, when the low-temperature frost damage prediction model has errors in predicting low-temperature frost damage, the critical antifreeze temperature threshold of wheat on a certain day, the data used to calculate the cold resistance training rate and the data used to calculate the cold resistance training release rate are the data with problems.

[0124] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A method for evaluating winter wheat frost injury mortality during winter, characterized in that: include: The first stage and the second stage of cold-resistance training of wheat are simulated respectively to obtain the first stage cold-resistance factor and the second stage cold-resistance factor; If the temperature of the wheat tillering node exceeds a preset threshold, the critical antifreeze temperature of the wheat that has not been hardened against cold is detected to determine the cold hardening release rate in combination with the critical antifreeze temperature of the wheat that has been hardened against cold; Obtaining the critical antifreeze temperature of the wheat one day before the evaluation; Constructing a low temperature frost damage prediction model according to the first-stage cold resistance factor, the second-stage cold resistance factor, the cold resistance training release rate, and the critical cold resistance temperature on the day before the evaluation; Calculate the observed value, simulated value, average value of the observed value, average value of the simulated value and the number of samples of the low temperature frost damage prediction model to determine the prediction accuracy, root mean square error, true deviation and significant difference of the low temperature frost damage prediction model; Determining the simulation accuracy of the cold hardship release effect according to the true deviation to determine whether to adjust the number of segments of the vernalization temperature or to adjust the threshold of the tillering node temperature of the cold hardship release effect according to the average difference between the critical cold hardiness temperatures of the root and stem of the wheat and the leaves after thawing; If the simulation accuracy still does not meet the requirements after the cold-resistant training coping method is executed, the ratio of data training to data replacement is adjusted according to the EF value of the low-temperature frost damage prediction model.

2. The method for evaluating winter wheat frost injury mortality rate according to claim 1, characterized in that: The calculation formula of the low temperature frost damage prediction model is: LT i =LT i-1 +f hardningi +f dehardningi Among them, LT i is the critical antifreeze temperature threshold on the i-th day during the wintering process, LT i-1 It is the critical antifreeze temperature of the variety the day before. hardningi is the cold resistance training rate on the i-th day, f dehardningi is the rate of cold resistance training release on day i; The lowest air temperature of the day is T min <LT i , it is determined that wheat has frost damage, when T min >LT i No frost damage occurs.

3. The method for evaluating winter wheat frost injury mortality rate during winter according to claim 2, characterized in that: The calculation formula of the first stage cold resistance factor is: f hardningi1 =f la ×ΔPot h1 ×min(f ti1 ,f shi ) Among them, f hardningi1 is the change in wheat temperature in the first stage of cold hardiness training, f ti1 is the temperature effect factor on the ith day in the first stage of cold-resistance training, f shi is the light effect factor on the i-th day, f la is the pre-winter leaf age factor, with a value between 0 and 1, △Pot h1 is the potential increase in daily wheat temperature after cold hardiness.

4. The method for evaluating winter wheat frost injury mortality rate during winter according to claim 3, characterized in that: The calculation formula of the cold resistance factor in the second stage is: Among them, f hardningi2 is the change of wheat temperature in the second stage of cold hardening, HF2 is the cold resistance effect factor in the first stage, R c1 R is the critical temperature of wheat frost resistance after the first stage of cold resistance training. c2 LT is the critical temperature of wheat frost resistance after the second stage of cold resistance training. i is the critical antifreeze temperature threshold on the i-th day during the wintering process, L max It is the critical antifreeze temperature of wheat that has been hardened to cold.

5. The method for evaluating winter wheat frost injury mortality rate during winter according to claim 4, characterized in that: The calculation formula for the prediction accuracy of the low temperature frost damage prediction model is: Among them, EF>0 means that the prediction accuracy of the low temperature damage prediction model meets the requirements; The root mean square error of the low temperature damage prediction model is the absolute error between the simulated value and the measured value, and the calculation formula is: Among them, RMSE a is the root mean square error of the low temperature damage prediction model; The true deviation of the low temperature damage prediction model is the sum of the deviations between the simulated value and the measured value, and the calculation formula is: Among them, R 2 is the true deviation of the low temperature damage prediction model; The significant difference of the low temperature damage prediction model is the difference between the simulated value and the measured value, and the calculation formula is: ABS=S i -O i ; Among them, ABS is the significant difference of low temperature damage prediction model; Among them, O i is the observed value, S i is the analog value, is the mean of the observed values, is the average value of the simulation, and n is the number of samples.

6. The method for evaluating winter wheat frost injury mortality rate during winter according to claim 5, characterized in that: Whether the simulation accuracy of the cold-hardiness training release effect meets the requirements is determined according to the actual deviation of the low-temperature frost damage prediction model. When the actual deviation of the low-temperature frost damage prediction model is greater than the preset second deviation, it is determined that the simulation accuracy of the cold-hardiness training release effect does not meet the requirements, and the number of segments of the vernalization temperature is increased.

7. The method for evaluating winter wheat frost injury mortality rate during winter according to claim 6, characterized in that: When the actual deviation of the low temperature frost damage prediction model is greater than the preset first deviation and less than or equal to the preset second deviation, it is determined that the simulation accuracy of the cold-hardening release effect does not meet the requirements, and a secondary judgment is made on the simulation accuracy of the cold-hardening release effect based on the average difference in the critical cold-hardiness temperature of the rhizomes and leaves of the wheat after thawing. When the average difference in the critical cold-hardiness temperature of the rhizomes and leaves of the wheat after thawing is greater than the preset temperature difference, the tillering node threshold of the cold-hardening release effect is adjusted.

8. The method for evaluating winter wheat frost injury mortality rate during winter according to claim 7, characterized in that: If the actual deviation of the low-temperature frost damage prediction model is still greater than the preset second deviation after adjustment, the simulation accuracy of the cold-resistant training release effect is determined according to the EF value of the low-temperature frost damage prediction model. If the EF value of the low-temperature frost damage prediction model is less than or equal to the EF value of the low-temperature frost damage model, it is determined that the simulation accuracy of the cold-resistant training release effect does not meet the requirements, and the ratio of data training to data replacement is adjusted.

9. The method for evaluating winter wheat frost injury mortality rate during winter according to claim 8, characterized in that: The average difference in critical cold-resistance temperatures of the rhizomes and leaves of the thawed wheat is the difference between the critical temperatures at which the rhizomes and leaves of the thawed wheat can withstand low temperatures without suffering frost damage; the wheat tillering node temperature refers to the extreme temperature condition that affects the growth and development of the wheat tillering node.

10. The method for evaluating winter wheat overwintering frost injury mortality according to claim 9, characterized in that: The ratio of data training to data replacement is the ratio of the amount of data used for training optimization problems in the data with problems when the low temperature damage prediction model is running to the amount of data used for search and replacement in the database; Among them, the sum of the data volume for the training optimization problem and the data volume for the search and replacement in the database is the data volume of data with problems when the low-temperature frost damage prediction model is running.

Images