Crop harvesting time determination method and system, server and medium

By inputting meteorological factors into deep learning models, predicting the quality and yield index values ​​of crops, determining the crop harvesting period, solving the quality and yield problems caused by farmers' premature picking, realizing scientific harvesting period decisions, and improving the quality of agricultural products and yield.

CN119026970APending Publication Date: 2024-11-26Chongqing Jiangjin District Meteorological Bureau
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410879385.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-02
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

Farmers pick crops too early and affect the quality and yield of agricultural products, making it difficult to choose the appropriate picking period.

Method used

By obtaining the meteorological factors during the predicted time period and inputting them into the deep learning model, the quality index value and yield index value are calculated, and the harvesting period of crops is determined based on these index values. Specific steps include screening meteorological factors with significant impact, establishing a random forest regression model, predicting crop quality and yield changes, and then determining the optimal harvesting period.

Benefits of technology

It effectively improves the quality and yield of agricultural products, predicts the quality and yield changes of crops through meteorological conditions, provides scientific guidance on the harvesting period, reduces the calculation process, and improves the calculation speed.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119026970A_ABST
    Figure CN119026970A_ABST
Patent Text Reader

Abstract

The invention provides a crop harvesting time determination method and system, a server and a medium, and the method comprises the steps: obtaining a meteorological factor of a prediction node in a prediction time period, and the prediction time period comprises a plurality of prediction nodes; inputting the meteorological factors into a deep learning model to obtain a quality index value and / or a yield index value; and determining a crop harvesting period based on the quality index value and / or the yield index value. According to the method, the harvesting time can be determined according to the emphasis on the actual requirements for quality and yield, and the quality of agricultural products can be improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of machine learning technology, and in particular to a method, system, server and medium for determining a harvest period of crops. Background Art

[0002] In recent years, with the rapid development of the deep processing industry, the development and processing technology of agricultural products has been greatly improved. In actual production, affected by market demand and cultivation technology, there is a problem of farmers picking agricultural products too early, which seriously affects the quality and yield of agricultural products. Therefore, choosing the right picking period is extremely important for the quality control of agricultural products. Summary of the invention

[0003] In view of this, one of the purposes of the embodiments of the present application is to provide a method for determining the harvest period of crops, which can improve the problem that premature picking of crops affects the quality and yield of agricultural products.

[0004] In order to achieve the above technical objectives, the technical solutions adopted in this application are as follows:

[0005] In a first aspect, an embodiment of the present application provides a method for determining a crop harvest period, the method comprising:

[0006] Obtaining meteorological factors of prediction nodes within a prediction time period, wherein the prediction time period includes a plurality of prediction nodes;

[0007] Inputting the meteorological factors into the deep learning model to obtain a quality index value and / or a yield index value, wherein the quality index value is used to characterize the quality of the crop at the corresponding prediction node, and the yield index value is used to characterize the yield of the crop at the corresponding prediction node;

[0008] The crop harvesting period is determined based on the quality index value and / or the yield index value.

[0009] Furthermore, the quality index value and / or yield index value is obtained by: inputting a meteorological factor whose correlation with the quality of the crop is greater than a first determination coefficient threshold into the deep learning model; and / or, inputting a meteorological factor whose correlation with the yield of the crop is greater than a second determination coefficient threshold into the deep learning model.

[0010] Furthermore, the method of determining the crop harvesting period is as follows: based on the quality index values ​​and / or yield index values ​​of all the prediction nodes, obtaining the change in quality grade and / or yield grade of the crop within the prediction time period, and determining the crop harvesting period according to the change in quality grade and / or yield grade.

[0011] Further, the change in quality grade and / or yield grade of the crop within the forecast time period is obtained as follows:

[0012] The predicted value of the quality grade or yield grade of the crop at each prediction node is calculated based on the following formula:

[0013]

[0014] X max X is the value of the crop just reaching the quality grade or the best quality grade in the yield grade. min is the value of the crop just reaching the quality grade or the good quality grade in the yield grade, T is the actual quality content value or yield value of the crop, obtained according to the quality index value or yield index value;

[0015] According to the time sequence, the change of the quality grade prediction value or the output grade prediction value of the prediction node is obtained.

[0016] Furthermore, the crop is Zanthoxylum bungeanum, the quality index value includes volatile oil content and numbing agent content, and the yield index value includes grain size and thousand-grain weight;

[0017] Before calculating the quality grade prediction value or yield grade prediction value of each prediction node, the following steps are performed to obtain the actual quality content value or yield value of the crop, the steps are:

[0018] The volatile oil content and the numbing hormone content are normalized, the particle size and the thousand-grain weight are normalized, or the volatile oil content, the numbing hormone content, the particle size and the thousand-grain weight are normalized.

[0019] Furthermore, the meteorological factor of the prediction node is obtained based on the meteorological factors of time nodes in the same period as the prediction node in previous years.

[0020] Furthermore, the deep learning model is a random forest regression model.

[0021] In a second aspect, the present application also provides a system for determining a crop harvest period, comprising:

[0022] An acquisition module configured to obtain meteorological factors of a prediction node within a prediction time period, wherein the prediction time period includes a plurality of prediction nodes;

[0023] A calculation module is configured to input the meteorological factors into a deep learning model to obtain a quality index value and / or a yield index value of the prediction node, wherein the quality index value is used to characterize the quality of the crop at the corresponding prediction node, and the yield index value is used to characterize the yield of the crop at the corresponding prediction node;

[0024] The prediction module is configured to determine the crop harvest period based on the quality index value and / or the yield index value.

[0025] In a third aspect, this embodiment further provides a server, comprising a processor and a memory coupled to each other, wherein the memory stores a computer program, and when the computer program is executed by the processor, the server executes the above-mentioned method for determining the crop harvest period.

[0026] In a fourth aspect, this embodiment further provides a computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium. When the computer program is executed on a computer, the computer executes the above-mentioned method for determining the crop harvest period.

[0027] The invention adopting the above technical solution has the following advantages:

[0028] In the technical solution provided by the present application, by inputting the meteorological factors of the prediction nodes into the deep learning model, the quality index value and / or yield index value of each prediction node is obtained, and the crop harvesting period is determined according to the quality index value and / or yield index value. That is, the present application obtains the quality index value and / or yield index value based on meteorological conditions, and then obtains the changes in the quality and / or yield of agricultural products over time. The harvesting period can be determined according to the actual demand for quality and yield, which is conducive to improving the quality of agricultural products.

[0029] In the technical solution provided in the present application, meteorological factors whose determination coefficients with the quality of crops are greater than a first determination coefficient threshold, and meteorological factors whose determination coefficients with the yield of crops are greater than a second determination coefficient threshold are input into the deep learning model to obtain quality index values ​​and yield index values, thereby reducing the calculation process and improving the calculation speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The present application may be further described by the non-limiting embodiments given in the accompanying drawings. It should be understood that the following drawings only illustrate certain embodiments of the present application and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, other relevant drawings may be obtained based on these drawings without creative effort.

[0031] Figure 1 A flow chart of the method provided in an embodiment of the present application.

[0032] Figure 2 Comparison of the importance of volatile oils and meteorological factors.

[0033] Figure 3 Comparison of volatile oil and meteorological factor training data sets.

[0034] Figure 4 Comparison of the importance of volatile oils and meteorological factors.

[0035] Figure 5 Comparison of volatile oil and meteorological factor training data sets.

[0036] Figure 6 Comparison of the importance of hemp flavor and meteorological factors.

[0037] Figure 7 Comparison of hemp flavor and meteorological factor training data sets.

[0038] Figure 8 Comparison of the importance of hemp flavor and meteorological factors.

[0039] Fig. 9 Comparison of hemp flavor and meteorological factor training data sets.

[0040] Fig.10 Comparison of the importance of thousand-grain weight and meteorological factors.

[0041] Fig.11 Comparison of thousand-grain weight and meteorological factor training data sets.

[0042] Fig.12 Comparison of the importance of thousand-grain weight and meteorological factors.

[0043] Fig.13 Comparison of thousand-grain weight and meteorological factor training data sets.

[0044] Fig.14 Comparison of the importance of particle size and meteorological factors.

[0045] Fig.15 Comparison of particle size and meteorological factor training data set.

[0046] Fig.16 Comparison of the importance of particle size and meteorological factors.

[0047] Fig.17 Comparison of particle size and meteorological factor training data set.

[0048] Fig.18 This is a chart showing the daily predicted changes in the quality grade scores of Sichuan pepper.

[0049] Fig.19 This is a chart showing the daily predicted changes in pepper yield grade scores.

[0050] Fig. 20 This is a chart showing the daily predicted changes in the suitability scores for picking pepper.

[0051] Fig.21 A structural diagram of the system provided in an embodiment of the present application.

[0052] Icons: 1-acquisition module; 2-calculation module; 3-prediction module. DETAILED DESCRIPTION

[0053] The present application will be described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that in the drawings or descriptions, similar or identical parts use the same figure numbers, and the implementation methods not shown or described in the drawings are forms known to ordinary technicians in the relevant technical field. In the description of this application, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0054] Please refer to Figure 1 The present application provides a method for determining a crop harvest period. The method for determining a crop harvest period may include the following steps:

[0055] S110: obtaining meteorological factors of prediction nodes within a prediction time period, where the prediction time period includes a number of prediction nodes;

[0056] S210: Inputting meteorological factors into the deep learning model to obtain quality index values ​​and / or yield index values;

[0057] S310: Determine the crop harvesting period based on the quality index value and / or the yield index value.

[0058] Taking pepper as an example, the steps of the method for determining the harvest period of crops are described in detail as follows:

[0059] In step 110, the predicted time period refers to the time period of interest. For example, in a certain place, the general picking time of peppercorns is from May to August, fresh peppercorns are mainly concentrated in June, and dried peppercorns are suitable for picking in July. Therefore, the predicted time period for dried peppercorns is July, and the predicted time period for fresh peppercorns is June. Through S110-S310, the harvest period of the corresponding peppercorn type can be obtained.

[0060] A prediction node represents a discrete point within a prediction time period. The length between adjacent prediction nodes represents a time unit. For example, if adjacent prediction nodes are one day, then a single prediction node is a specific time of the day, such as April 28, May 31, etc. The time unit is 1 day, or it can be 3 days, 5 days, etc. A prediction time period includes several prediction nodes.

[0061] Meteorological factors represent meteorological causes or conditions that affect the development and change of other things. Each prediction node corresponds to several types of meteorological factors. It can be understood that when the prediction node represents a natural day, the meteorological factor can represent the meteorological conditions of the corresponding natural day.

[0062] In S210, the types of meteorological factors input into the deep model should be screened, and meteorological factors with significant influence are required, otherwise the computing power requirements of the deep model will increase. The quality index values ​​of pepper include volatile oil content and numbing substance content, and the yield index values ​​include grain size and thousand-grain weight. The screening of meteorological factors is achieved by the following method:

[0063] The meteorological factors whose determination coefficients with the quality of Sichuan pepper are greater than the first determination coefficient threshold are input into the deep learning model; and / or, the meteorological factors whose determination coefficients with the yield of Sichuan pepper are greater than the second determination coefficient threshold are input into the deep learning model.

[0064] There may be multiple types of meteorological factors that are strongly correlated with the quality and yield of pepper, so in this embodiment, the specific value of the first determination coefficient and the specific value of the second determination coefficient can be flexibly set according to the type of meteorological factor.

[0065] In this embodiment, the quality index values ​​of Zanthoxylum bungeanum include volatile oil content and numbing element content, and the yield index values ​​of Zanthoxylum bungeanum include grain size and thousand-grain weight. By establishing a random forest regression model, meteorological factors with greater correlation with quality and yield are selected.

[0066] Specifically, the screening method of meteorological factors affecting the volatile oil content is as follows:

[0067] The meteorological factors that have a significant impact on volatile oils include the 40-day average daily range, the 35-day average sunshine hours, the 60-day average precipitation, and the 40-day average high humidity days to establish the model. Figure 2 It can be seen that the importance of 35-day sunshine hours in the model is 0.1, which is relatively low compared to other meteorological factors. Figure 3 It can be seen that the determination coefficient R can be seen from the training set and the actual situation 2 The model is 0.588, and the fitting degree of the model is general. In order to further improve the performance of the model, the meteorological factors are simplified, the factors with higher importance are selected to establish the model again, and the model is re-established for the relationship between the quality of volatile oil and the daily difference of 40 days, the precipitation of 60 days, and the number of days with high humidity of 40 days. Figure 4 and Figure 5 It can be seen that the 60-day precipitation has a greater impact on the volatile oil of Zanthoxylum bungeanum. The effect of the model has been improved after adjustment. The determination coefficient R 2 Increased to 0.65.

[0068] Table 1 Meteorological factors corresponding to different grades of volatile oil

[0069]

[0070] The random forest regression model was used to calculate the prediction of volatile oil under different meteorological factors, and Table 1 was obtained. It can be seen from Table 1 that when the 40-day average daily range is larger, the quality of the volatile oil will decrease, but the quality can only be better when the temperature is above 4°C; the greater the 60-day average precipitation, the better the impact on the quality of the volatile oil; the more 40-day average high humidity days, the better the impact on the quality of the volatile oil.

[0071] The screening method of meteorological factors affecting hemp flavor is as follows:

[0072] The meteorological factors that have a significant impact on pepper and numbing essence are selected, including the 30-day average daily temperature difference, the 55-day average minimum temperature, the 25-day average sunshine hours, the 30-day average relative humidity days, and the 45-day average precipitation. Figure 6 , Figure 7 ,from Figure 7 The coefficient of determination R of the model is shown 2 is 0.795, the model effect is good. Figure 6 It can be seen that the 45-day average precipitation has a low importance. Removing it and remodeling can obtain Figure 8 , Fig. 9 , the coefficient of determination R 2 is 0.8, the model works well.

[0073] Table 2 Meteorological factors corresponding to different levels of numbing agents

[0074]

[0075]

[0076] The random forest regression model was used to calculate the prediction of Zanthoxylum bungeanum under different meteorological factors, and Table 2 was obtained. It can be seen from Table 2 that the 30-day average daily temperature difference has the greatest impact on Zanthoxylum bungeanum. The smaller the daily temperature difference, the better the quality of Zanthoxylum bungeanum. The lower the average minimum temperature, the lower the sunshine hours, and the average relative humidity maintained in the range of 60-90% have the greatest impact on the quality of Zanthoxylum bungeanum.

[0077] The screening method of meteorological factors affecting thousand-grain weight is as follows:

[0078] The meteorological factors that have a significant impact on the thousand-grain weight of pepper include the 50-day average precipitation, the 45-day average temperature, the 45-day average maximum temperature, and the 50-day average minimum temperature to establish a model. Fig.10 , Fig.11 , it can be seen that the importance of the 50-day average precipitation is relatively low, and the importance of the model is not high, while the model determination coefficient R 2 To improve the accuracy of the model, the 50-day average precipitation data was deleted and the model was rebuilt. Fig.12 , Fig.13 , we can see that the new model determination coefficient R 2 It is increased to 0.75, and the prediction effect of the model is greatly improved.

[0079] Table 3 Meteorological factors corresponding to different grades of thousand-grain weight

[0080]

[0081] The random forest regression model was used to calculate the prediction of thousand-grain weight under different meteorological factors, and Table 3 was obtained. It can be seen from Table 3 that the important meteorological factors affecting thousand-grain weight are all related to temperature. The higher the 45-day average temperature, the 45-day average maximum temperature and the 50-day average minimum temperature, the larger the thousand-grain weight and the better the yield performance.

[0082] The screening method of meteorological factors affecting particle size is as follows:

[0083] The meteorological factors that have a significant impact on the grain size of pepper include 50-day average precipitation, 45-day average temperature, 25-day average minimum temperature, 10-day average minimum temperature, and 35-day average maximum temperature to establish a model. Fig.14 , Fig.15 It can be seen that the importance of each meteorological factor in the model is relatively low in terms of the 45-day average temperature and the 10-day average minimum temperature. By reselecting the meteorological factors to establish the model, we can get Fig.16 , Fig.17 , model determination coefficient R 2 It is 0.85, which has a good effect.

[0084] Table 4 Meteorological factors corresponding to different particle size levels

[0085]

[0086]

[0087] The random forest regression model was used to calculate the prediction of particle size under different meteorological factors, and Table 4 was obtained. It can be seen from Table 4 that the particle size is mainly affected by the temperature. The higher the temperature, the larger the particle size. To achieve excellent quality, the average maximum temperature in 35 days must be above 26°C, and the average temperature in 45 days must be above 22°C.

[0088] Therefore, the meteorological factors affecting the volatile oil content are the 40-day average daily temperature range, the 60-day average precipitation and the 40-day average high humidity; the meteorological factors affecting the numbing agent are the 30-day average daily temperature range, the 55-day average minimum temperature, the 25-day average sunshine hours and the 30-day average relative humidity; the meteorological factors affecting the 1000-grain weight are the 45-day average temperature, the 45-day average maximum temperature and the 50-day average minimum temperature; the meteorological factors affecting the particle size are the 35-day average maximum temperature, the 25-day average minimum temperature and the 45-day average temperature.

[0089] The above meteorological factors are input into the deep learning model. The deep learning model in this embodiment is a random forest regression model. The training method of the random forest regression model is:

[0090] 1. Select data from 2019-2022 as the data set and divide the data set into a training set and a test set. Usually 80% of the data is used as the training set and the remaining 20% ​​is used as the test set. The data set includes the volatile oil content, sesame content, thousand-grain weight, particle size and meteorological factors at each time point from 2019 to 2022.

[0091] 2. Build a random forest regression model: Use the random forest regression model in the machine learning library, set the parameters of the random forest for different qualities, and adjust the number of trees (n_estimators), maximum depth (max_depth), and minimum number of sample splits (min_samples_split).

[0092] 3. Model training: Use the training set data to train the random forest regression model and monitor the performance indicators during the training process (such as mean square error MSE, determination coefficient R 2 etc.) in order to adjust parameters or perform other optimizations in a timely manner, and continuously optimize the random forest regression model through the test machine.

[0093] 4. Model deployment and monitoring: Deploy the trained model to the actual production environment, and predict the changes in pepper quality and yield by setting meteorological factors related to temperature, precipitation, and humidity. The performance of the model still needs to be monitored regularly, and the model needs to be retrained or adjusted when necessary to adapt to new data or environmental changes.

[0094] The aforementioned random forest regression model can be used to implement the aforementioned screening of meteorological factors that have a greater impact on volatile oil content, sesame content, thousand-grain weight and particle size.

[0095] In S210, the screened meteorological factors of each prediction node are input into the above-mentioned random forest regression model to obtain the quality index value (volatile oil content and sesame content) and yield index value (thousand-grain weight and particle size) of each prediction node.

[0096] In this embodiment, the meteorological factors of the prediction node are selected based on the meteorological factors of the time nodes of the same period as the prediction node in previous years. For example, if the applicant wants to obtain the volatile oil content in the next month (assuming it is from X month 1st to X month 31st, and the distance between adjacent prediction nodes is 1 day), and determine the harvest period only based on the volatile oil content, the 40-day average daily range, 60-day average precipitation and 40-day average high humidity days for each day from X month 1st to X month 31st in the past ten years can be collected, and after being processed by a preset mathematical formula or mathematical model, they are input into the deep learning model.

[0097] Alternatively, the filtered meteorological factors of the prediction nodes obtained based on the meteorological forecast data are input into the deep learning model.

[0098] In S310, the method for determining the crop harvesting period is: based on the quality index values ​​and / or yield index values ​​of all prediction nodes, obtain the change in quality grade and / or yield grade of the crops within the prediction time period, and determine the crop harvesting period according to the change in quality grade and / or yield grade.

[0099] In this embodiment, the quality grade and yield grade are divided into three levels, namely, excellent, good and general. According to the quality index value and yield index value of each prediction node, the change of quality grade and yield grade in the prediction time period can be obtained, and then the crop harvest period can be determined according to the above changes. The values ​​defining excellent, good and general can be set according to the actual situation of the location of the pepper, for example:

[0100] A K-means cluster analysis was performed on the quality and yield index data of a major pepper producing area from 2016 to 2020. The number of clusters was set to 3, and the average values ​​of adjacent cluster centers of pepper quality indicators were used as boundaries. The evaluation standards of pepper quality and yield indicators were uniformly divided into 3 levels, namely, level 1 excellent, level 2 good, and level 3 general (Table 5).

[0101] Table 5 Green pepper quality and yield grade classification

[0102]

[0103] However, the units of volatile oil content, narcotic substance content, thousand-grain weight and particle size are different, which is not conducive to statistical analysis. In this embodiment, after calculating the volatile oil content, narcotic substance content, thousand-grain weight and particle size corresponding to each prediction node, the volatile oil content, narcotic substance content, thousand-grain weight and particle size are normalized to make the four become dimensionless data, and the normalized volatile oil content, narcotic substance content, thousand-grain weight and particle size are obtained.

[0104] Substitute the normalized volatile oil content, numbing hormone content, thousand-grain weight and particle size into the following formula respectively to calculate the predicted value of volatile oil content grade, numbing hormone content grade, thousand-grain weight grade and particle size grade. The average value of the predicted value of volatile oil content grade and numbing hormone content grade is taken as the predicted value of quality grade, and the average value of the predicted value of thousand-grain weight grade and particle size grade is taken as the predicted value of yield grade.

[0105]

[0106] X max X is the value of the quality grade or the best quality grade in the yield grade. min is the value just reaching the good quality grade in the quality grade or yield grade, and T is the actual quality content value or yield value, that is, the normalized volatile oil content, numbing hormone content, thousand-grain weight or particle size.

[0107] In this embodiment, the general grade is less than 60 points, the good grade is 60-80 points, and the excellent grade is above 80 points.

[0108] The following table is obtained after predicting the harvest period from May to August in a certain area.

[0109] Table 6 Green pepper quality and yield prediction

[0110]

[0111] The predicted value of picking suitability is obtained based on the average of the predicted values ​​of quality grade and yield grade.

[0112] From Table 6, we can get Fig.18 , Fig.19 and Fig. 20 .

[0113] from Fig.18 It can be seen that the overall quality of Sichuan pepper did not change much in the early stage. The quality grade score began to change from June 29th, and maintained at the highest from July 4th to 7th. The quality grade score then began to gradually decline. The quality began to decline significantly from August 13th, but it can be found that the quality score basically remained at around 70, and the picking period was in the good quality grade range. Therefore, if you only focus on the quality, you can pick it in a relatively high quality grade range from June 29th to August 13th.

[0114] from Fig.19It can be seen that the yield of Sichuan pepper changes significantly over time. Starting from May 18, the yield grade score became above 60, reaching the good grade level. On July 14, the yield grade score became above 80, reaching the excellent grade level. Subsequently, the yield grade did not change significantly. Therefore, if we only focus on the yield situation, it has been in an excellent quality grade range since May 15 and is suitable for picking.

[0115] See Table 6 and Fig. 20 It can be seen that in terms of quality, the quality of Sichuan pepper remained basically unchanged from May to June, and the value of quality change over time was not obvious. The quality of Sichuan pepper has been maintained at the good level for a long time, which is closely related to the accumulation of meteorological conditions in the previous period. The quality began to change significantly in July. In terms of yield, the yield of Sichuan pepper began to gradually increase from May. From May 19th, the yield grade reached the good level. It can be seen that the yield index is closely related to the increase in temperature. With the increase in temperature, it is expected that the yield can reach the excellent level on July 14th. Therefore, Sichuan pepper can be gradually picked around May 19th. If you want better yield and quality, you can start picking around July 14th.

[0116] Through the above method, the quality index value and yield index value of each prediction node can be obtained, that is, the volatile oil content, numbing substance content, thousand-grain weight and particle size. According to the changes in the quality index value and yield index value of all prediction nodes, the prediction of the suitable harvest period of pepper for different purposes can be realized. For example, the pepper used to make pepper sauce is suitable for harvesting from the end of April to the beginning of May. If it is picked before the end of April, the numbing substance content is low, resulting in insufficient numbing flavor, and the thousand-grain weight of pepper is low, which affects the yield. To produce pepper sauce with low numbing degree, it can be picked early within the suitable harvesting time range. The suitable starting period for preserving pepper is the end of early June. Picking pepper too early not only lacks numbing flavor, affecting the edible quality of fresh pepper, but also makes the pepper not full enough, and the low thousand-grain weight affects the pepper yield. It is not easy to pick dried peppercorns too early. It is best to start picking them in late June. On the one hand, they have higher volatile oil and numbing substances, and at the same time, the content of hydroxy-α-sanshool is at a higher level. Picking peppercorns at this time can ensure the fragrance and numbing taste of dried peppercorns. It is suitable to start picking peppercorns for oil production after late June to ensure that the peppercorn oil has a numbing aroma and low bitterness. The harvesting period should not last too long to ensure that the summer shoots emerge before the summer drought (before mid-July). In order not to affect the branching of peppercorns, the picking should be completed before mid-July. Harvesting can be appropriately delayed in places with water irrigation.

[0117] Please refer to Fig.21 This embodiment proposes a system for determining the harvest period of crops, based on the above method, including:

[0118] The acquisition module 1 is configured to obtain meteorological factors of prediction nodes within a prediction time period, and the prediction time period includes a plurality of prediction nodes.

[0119] The calculation module 2 is configured to input the meteorological factors into the deep learning model to obtain the quality index value and / or yield index value of the prediction node.

[0120] The prediction module 3 is configured to determine the crop harvest period based on the quality index value and / or the yield index value.

[0121] The embodiment of the present application provides a server that may include a processing module and a storage module. The storage module stores a computer program, and when the computer program is executed by the processing module, the server can execute the corresponding steps in the above-mentioned method for determining the harvest period of crops.

[0122] In this embodiment, the processing module may be an integrated circuit chip having the ability to process signals. The above-mentioned processing module may be a general-purpose processor. For example, the processor may be a central processing unit (CPU), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, and may implement or execute the disclosed methods, steps and logic block diagrams in the embodiments of the present application.

[0123] The storage module may be, but is not limited to, a random access memory, a read-only memory, a programmable read-only memory, an erasable programmable read-only memory, an electrically erasable programmable read-only memory, etc. In this embodiment, the storage module may be used to store preset numbers, etc. Of course, the storage module may also be used to store a program, and the processing module executes the program after receiving the execution instruction.

[0124] It should be noted that those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the server described above can refer to the corresponding process of each step in the aforementioned method, and will not be elaborated here.

[0125] The embodiment of the present application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed on a computer, the computer executes the method for determining the crop harvest period as described in the above embodiment.

[0126] Through the description of the above implementation methods, technical personnel in this field can clearly understand that the present application can be implemented by hardware, and can also be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present application.

[0127] In the embodiments provided in the present application, it should be understood that the disclosed system and method can also be implemented in other ways. The system and method embodiments described above are merely schematic, for example, the flow charts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the systems, methods and computer program products according to the multiple embodiments of the present application. In this regard, each box in the flow chart or block diagram can represent a part of a module, a program segment or a code, and a part of the module, program segment or code includes one or more executable instructions for implementing the specified logical function. It should also be noted that each box in the block diagram and / or the flow chart, and the combination of the boxes in the block diagram and / or the flow chart can be implemented by a dedicated hardware-based system that performs the specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions. In addition, each functional module in each embodiment of the present application can be integrated together to form an independent part, or each module can exist separately, or two or more modules can be integrated to form an independent part.

[0128] The above description is only an embodiment of the present application and is not intended to limit the protection scope of the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for determining the harvest period of crops, characterized in that: The method comprises: Obtaining meteorological factors of prediction nodes within a prediction time period, wherein the prediction time period includes a plurality of prediction nodes; Inputting the meteorological factors into the deep learning model to obtain a quality index value and / or a yield index value, wherein the quality index value is used to characterize the quality of the crop at the corresponding prediction node, and the yield index value is used to characterize the yield of the crop at the corresponding prediction node; The crop harvesting period is determined based on the quality index value and / or the yield index value.

2. The method according to claim 1, characterized in that The quality index value and / or yield index value is obtained by: inputting a meteorological factor whose correlation with the quality of the crop is greater than a first determination coefficient threshold into the deep learning model; and / or, inputting a meteorological factor whose correlation with the yield of the crop is greater than a second determination coefficient threshold into the deep learning model.

3. The method according to claim 1, characterized in that The method of determining the crop harvesting period is as follows: based on the quality index values ​​and / or yield index values ​​of all the prediction nodes, obtaining the change in quality grade and / or yield grade of the crop within the prediction time period, and determining the crop harvesting period according to the change in quality grade and / or yield grade.

4. The method according to claim 3, characterized in that The change in quality grade and / or yield grade of the crop within the forecast period is obtained as follows: The predicted value of the quality grade or yield grade of the crop at each prediction node is calculated based on the following formula: X max X is the value of the crop just reaching the quality grade or the best quality grade in the yield grade. min is the value of the crop just reaching the quality grade or the good quality grade in the yield grade, T is the actual quality content value or yield value of the crop, obtained according to the quality index value or yield index value; According to the time sequence, the change of the quality grade prediction value or the output grade prediction value of the prediction node is obtained.

5. The method according to claim 4, characterized in that The crop is Zanthoxylum bungeanum, the quality index values ​​include volatile oil content and numbing agent content, and the yield index values ​​include grain size and thousand-grain weight; Before calculating the quality grade prediction value or yield grade prediction value of each prediction node, the following steps are performed to obtain the actual quality content value or yield value of the crop, the steps are: The volatile oil content and the numbing hormone content are normalized, the particle size and the thousand-grain weight are normalized, or the volatile oil content, the numbing hormone content, the particle size and the thousand-grain weight are normalized.

6. The method according to claim 1, characterized in that The meteorological factor of the prediction node is obtained based on the meteorological factors of time nodes in the same period as the prediction node in previous years.

7. The method according to claim 1, characterized in that The deep learning model is a random forest regression model.

8. A system for determining the harvesting period of crops, characterized in that: include: An acquisition module configured to obtain meteorological factors of a prediction node within a prediction time period, wherein the prediction time period includes a plurality of prediction nodes; A calculation module is configured to input the meteorological factors into a deep learning model to obtain a quality index value and / or a yield index value of the prediction node, wherein the quality index value is used to characterize the quality of the crop at the corresponding prediction node, and the yield index value is used to characterize the yield of the crop at the corresponding prediction node; The prediction module is configured to determine the crop harvest period based on the quality index value and / or the yield index value.

9. A server, characterized in that: The server includes a processor and a memory coupled to each other, wherein the memory stores a computer program. When the computer program is executed by the processor, the server executes the method for determining the crop harvest period as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed on a computer, the computer is enabled to execute the method for determining the harvest period of crops according to any one of claims 1 to 7.