Irrigation and fertilization optimization method and device for corn and soybean intercropping and storage medium

Abstract

The invention discloses an irrigation and fertilization optimization method and device for corn and soybean intercropping and a storage medium. Relates to the technical field of fertigation optimization. The method comprises the following steps: acquiring historical illumination intensity information of a first area at a current time point; soybeans are planted in the first area, and corn is planted in the second area close to the first area; the first area comprises at least two sub-areas which are sequentially arranged in the direction away from the second area; the historical illumination intensity information comprises historical illumination intensity information of all sub-regions; according to the historical illumination intensity information, utilizing a neural network model to determine illumination intensity prediction information of each sub-region within a preset time after the current time point; and determining the fertigation scheme of each sub-region at the current time point from the fertigation scheme set according to the illumination intensity prediction information. The technical problem that the yield of the intercropped soybeans is seriously affected due to the fact that the shading effect of the corn on the soybeans is not considered in an irrigation and fertilization scheme in a corn and soybean intercropping mode is solved.

Description

Technical Field

[0001] The present application relates to the technical field of irrigation and fertilization optimization, and in particular to an irrigation and fertilization optimization method, device and storage medium for corn and soybean intercropping. Background Art

[0002] In the soybean-corn intercropping model, soybeans, as a low-lying crop, are easily affected by the shading of the neighboring high-lying corn crop. This shading effect intensifies as corn enters the jointing stage, manifested as rapid growth of corn plants, which can reach a height of 250-300 cm, significantly blocking the light of soybeans, making the soybeans next to the corn rows the most shaded. In this case, the utilization rate of light energy of soybeans decreases, and coupled with the competition of corn for water and nutrients, the soybean stems become thinner and the plant height is higher than that of single-cropping, and the lodging rate is greatly increased. Irrational fertilization will further lead to the lodging of soybeans, which seriously affects the yield and quality of crops.

[0003] The current fertigation practice only considers that soybeans can meet half of their nitrogen needs through biological nitrogen fixation, so soybeans do not need a large amount of nitrogen fertilizer. However, if the nitrogen fertilizer supply of corn is insufficient, it will cause a reduction in yield. If fertilizer is applied uniformly to the entire field, it will lead to insufficient fertilizer supply for corn or excessive growth and lodging of soybeans. Therefore, it is chosen to fertilize the two crops separately according to the crop type. However, this fertigation method only applies fertilizer universally according to the crop type, fails to fully consider the differences in soybean growth environment caused by corn shading, and cannot cope with the shading effect of corn on soybeans, which seriously affects the yield of intercropped soybeans.

[0004] With regard to the technical problem that the irrigation and fertilization scheme in the corn-soybean intercropping mode in the above-mentioned prior art does not take into account the shading effect of corn on soybeans, which seriously affects the yield of intercropped soybeans, no effective solution has been proposed so far. Summary of the invention

[0005] The embodiments of the present disclosure provide a method, device and storage medium for optimizing fertilization and irrigation in corn and soybean intercropping, so as to at least solve the technical problem that the fertilization and irrigation scheme in the corn and soybean intercropping mode in the prior art does not consider the shading effect of corn on soybeans, which seriously affects the yield of intercropped soybeans.

[0006] According to one aspect of an embodiment of the present disclosure, there is provided a method for optimizing irrigation and fertilization for intercropping of corn and soybeans, comprising: obtaining historical light intensity information of a first area at a current time point; wherein soybeans are planted in the first area, and corn is planted in a second area adjacent to the first area; the first area comprises at least two sub-areas arranged in sequence in a direction away from the second area; the historical light intensity information comprises historical light intensity information of all sub-areas; based on the historical light intensity information, determining light intensity prediction information of each sub-area within a preset time after the current time point using a pre-trained neural network model; wherein the preset time is the next fertilization cycle of the soybeans; and determining, based on the light intensity prediction information, an irrigation and fertilization scheme for each sub-area at the current time point from a preset set of irrigation and fertilization schemes.

[0007] According to another aspect of the embodiments of the present disclosure, a storage medium is further provided. The storage medium includes a stored program, wherein the above method is executed by a processor when the program is running.

[0008] According to another aspect of an embodiment of the present disclosure, there is also provided an irrigation and fertilization optimization device for intercropping of corn and soybeans, comprising: an information acquisition module, for acquiring historical light intensity information of a first area at a current time point; wherein the first area is planted with soybeans, and a second area adjacent to the first area is planted with corn; the first area comprises at least two sub-areas arranged in sequence in a direction away from the second area; the historical light intensity information comprises historical light intensity information of all sub-areas; an information determination module, for determining, based on the historical light intensity information, light intensity prediction information of each sub-area within a preset time after the current time point using a pre-trained neural network model; wherein the preset time is the next fertilization cycle of the soybeans; and a scheme determination module, for determining, based on the light intensity prediction information, an irrigation and fertilization scheme for each sub-area at the current time point from a preset set of irrigation and fertilization schemes.

[0009] According to another aspect of an embodiment of the present disclosure, there is also provided an irrigation and fertilization optimization system for intercropping of corn and soybeans, comprising a processor; and a memory connected to the processor, for providing the processor with instructions for processing the following processing steps: at a current time point, obtaining historical light intensity information of a first area; wherein the first area is planted with soybeans, and a second area adjacent to the first area is planted with corn; the first area includes at least two sub-areas arranged in sequence in a direction away from the second area; the historical light intensity information includes historical light intensity information of all sub-areas; based on the historical light intensity information, determining light intensity prediction information of each sub-area within a preset time after the current time point using a pre-trained neural network model; wherein the preset time is the next fertilization cycle of the soybeans; and based on the light intensity prediction information, determining an irrigation and fertilization scheme for each sub-area at the current time point from a preset set of irrigation and fertilization schemes.

[0010] The present application takes into account that there is a clear difference in the shading effect between the position near the corn and the position far from the corn in the first area, and the first area is divided into at least two sub-areas arranged in sequence along the direction away from the second area. When it is necessary to irrigate and fertilize the intercropped soybeans at the current time point, the historical light intensity information of all sub-areas in the first area is first obtained to provide data support for the subsequent prediction of light intensity. Then, considering the different degrees of influence of corn shading on different sub-areas and the differences in the fertilization cycles of soybeans at different growth stages, according to these historical light intensity information, the light intensity prediction information of each sub-area in the next fertilization cycle of soybeans after the current time point is determined using a pre-trained neural network model to achieve refined and differentiated dynamic prediction. Finally, according to the light intensity prediction information, the irrigation and fertilization scheme of each sub-area at the current time point is determined from a preset set of irrigation and fertilization schemes to achieve a personalized response to the growth needs of soybeans under different light conditions. Thus, the present application accurately considers the shading effect of corn on soybeans and achieves the purpose of refined irrigation and fertilization that is adapted to local conditions and time. This solves the technical problem that the irrigation and fertilization scheme in the corn-soybean intercropping mode in the prior art does not take into account the shading effect of corn on soybeans, which seriously affects the yield of intercropped soybeans. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The drawings described herein are used to provide a further understanding of the present disclosure and constitute a part of the present application. The illustrative embodiments of the present disclosure and their descriptions are used to explain the present disclosure and do not constitute an improper limitation on the present disclosure. In the drawings: Figure 1 is a hardware structure block diagram of a computing device for implementing the method according to Embodiment 1 of the present application; Figure 2aIt is a schematic diagram of various areas intercropped with corn and soybeans according to Example 1 of the present application; Figure 2b is another schematic diagram of various regions intercropped with corn and soybeans according to Example 1 of the present application; Figure 3 It is a schematic diagram of the process of optimizing the irrigation and fertilization method for intercropping corn and soybeans according to Example 1 of the present application; Figure 4 is a schematic diagram of the irrigation and fertilization optimization device for corn and soybean intercropping according to Example 2 of the present application; and Figure 5 It is a schematic diagram of the irrigation and fertilization optimization system for corn and soybean intercropping described in Example 3 of the present application. DETAILED DESCRIPTION

[0012] In order to enable those skilled in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only embodiments of a part of the present disclosure, not all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of the present disclosure.

[0013] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products, or devices.

[0014] Example 1 According to this embodiment, a method embodiment of an irrigation and fertilization optimization method for intercropping corn and soybeans is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0015] The method embodiment provided in this embodiment can be executed in a mobile terminal, a computer terminal, a server or a similar computing device. Figure 1 FIG. 1 shows a hardware structure block diagram of a computing device for implementing an irrigation and fertilization optimization method for intercropping corn and soybeans. Figure 1 As shown, the computing device may include one or more processors (the processor may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory for storing data, a transmission device for communication functions, and an input / output interface. The memory, the transmission device, and the input / output interface are connected to the processor via a bus. In addition, it may also include: a display, a keyboard, and a cursor control device connected to the input / output interface. A person skilled in the art can understand that Figure 1 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.

[0016] It should be noted that the one or more processors and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuits may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the computing device. As involved in the embodiments of the present disclosure, the data processing circuit acts as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0017] The memory can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the irrigation and fertilization optimization method for corn and soybean intercropping in the embodiment of the present disclosure. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, the irrigation and fertilization optimization method for corn and soybean intercropping of the above-mentioned application program is realized. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the computing device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0018] The transmission device is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the computing device. In one example, the transmission device includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0019] The display may be, for example, a touch screen liquid crystal display (LCD) that may enable a user to interact with a user interface of the computing device.

[0020] It should be noted that, in some optional embodiments, the above Figure 1 The computing device shown may include hardware elements (including circuits), software elements (including computer code stored on a computer-readable medium), or a combination of both hardware elements and software elements. Figure 1 This is merely one example of a particular embodiment and is intended to illustrate the types of components that may be present in the computing devices described above.

[0021] Figure 2a Schematic diagram of various areas of intercropping with corn and soybean according to this embodiment. Figure 2a As shown in FIG. 1 , soybeans are planted in the first area, and corns are planted in the second area adjacent to the first area (which can be the left or right side). Considering that soybeans planted at positions adjacent to the second area in the first area are significantly affected by the shading of corns than those planted at other positions away from the second area, the first area can be divided into two sub-areas in a direction away from the second area, as shown in FIG. Figure 2a The first sub-region and the second sub-region are shown in the figure, wherein the first sub-region is closer to the second region than the first sub-region, that is, the soybeans planted in the first sub-region are greatly affected by the shading of the corn planted in the second region.

[0022] Figure 2b is another schematic diagram of various areas of intercropping with corn and soybean according to this embodiment. Figure 2b As shown in FIG. 1 , soybeans are planted in the first area, and corns are planted in the second area adjacent to the first area on both sides (including the left and right sides). Similarly, considering that soybeans planted in the first area adjacent to the second area are significantly affected by the shading of corns than those planted in other areas away from the second area, the first area can be divided into three sub-areas in the direction away from the second area, as shown in FIG. Figure 2bThere are two first sub-areas and one second sub-area shown in the figure, wherein the left first sub-area is closer to the second area arranged on the left side of the first area than the second sub-area, and the right first sub-area is closer to the second area arranged on the right side of the first area than the second sub-area, that is, the soybeans planted in these two first sub-areas are greatly affected by the shading of the corn planted in the second area.

[0023] Under the above operating environment, according to the first aspect of this embodiment, a method for optimizing irrigation and fertilization for intercropping of corn and soybean is provided. Figure 3 A schematic diagram showing the process of the method is shown in FIG. Figure 3 As shown, the method includes: S102: At the current time point, historical light intensity information of a first area is obtained; wherein the first area is planted with soybeans, and a second area adjacent to the first area is planted with corn; the first area includes at least two sub-areas arranged in sequence in a direction away from the second area; and the historical light intensity information includes historical light intensity information of all sub-areas; S104: Determine, based on the historical light intensity information, light intensity prediction information of each sub-region within a preset time after the current time point using a pre-trained neural network model; wherein the preset time is the next fertilization cycle of the soybean; S106: Determine, according to the light intensity prediction information, the fertilization and irrigation scheme for each sub-area at the current time point from a preset fertilization and irrigation scheme set.

[0024] Specifically, refer to Figure 2a and Figure 2b As shown, the first area specifically refers to the area where soybeans are planted, while the second area immediately outside the first area is planted with corn. Considering that there is a significant difference in the shading effect between the position close to the corn and the position far from the corn in the first area, the first area is further subdivided into a plurality of sub-areas arranged in sequence along the direction away from the second area according to the distance from the corn. This division is intended to capture the changes in lighting conditions caused by the different distances from the corn. For each sub-area, considering that the shadow effect of corn on soybeans changes with time and season, it is necessary to obtain the light intensity record of the sub-area in the past period of time at the current time point to determine the historical light intensity information of all sub-areas (corresponding to step 102) to provide data support for the subsequent prediction of light intensity.

[0025] Then, considering that different sub-regions are affected by corn shading to different degrees, based on the historical light intensity information of these sub-regions, the light intensity information of each sub-region in a specific time period in the future is predicted through a pre-trained neural network model (corresponding to step 104). Considering the differences in the fertilization cycles of soybeans at different growth stages, the specific time period in the future here refers to the period from the current time point to the end of the next soybean fertilization cycle, ensuring that the prediction results are directly related to the actual management decisions of the crops. Thus, the predicted light intensity value of each sub-region in a future fertilization cycle is determined by the neural network model. These predicted values ​​reflect the degree of influence of the corn shadow effect on the light reception of each sub-region, and realize refined and differentiated dynamic prediction.

[0026] Finally, according to the light intensity prediction information, the fertilization and irrigation scheme for each sub-area at the current time point is determined from the preset fertilization and irrigation scheme set (corresponding to step 106). The fertilization and irrigation scheme set includes various possible fertilization and irrigation schemes for different light conditions and soybean growth stages. The fertilization and irrigation scheme for each sub-area will be customized according to its specific light intensity prediction information, so that in the same field, different sub-areas should receive different water and fertilizer inputs according to their light conditions, so as to optimize the growth conditions of soybeans and ensure that the soybeans in each sub-area can receive the most suitable care for their light conditions and growth status, thereby achieving a personalized response to the growth needs of soybeans under different light conditions.

[0027] As described in the background technology, in the soybean-corn intercropping mode, soybeans, as low-lying crops, are easily affected by the shading of the neighboring high-lying crop corn, and the current fertigation practice only considers that soybeans can meet half of the nitrogen requirement of soybeans through biological nitrogen fixation, so soybeans do not need to invest a large amount of nitrogen fertilizer, and if the nitrogen fertilizer supply of corn is insufficient, it will cause a reduction in production. If fertilizer is applied uniformly throughout the field, it will lead to insufficient fertilizer supply for corn or lodging of soybeans due to vigorous growth, so it is chosen to fertilize the two crops separately according to the crop type. However, this fertigation method only applies fertilizer universally according to the crop type, fails to fully consider the differences in soybean growth environment caused by corn shading, and cannot cope with the shading effect of corn on soybeans, which seriously affects the yield of intercropped soybeans.

[0028] In view of this, the present application considers that there is a significant difference in the shading effect between the position near the corn and the position far from the corn in the first area, and the first area is divided into at least two sub-areas arranged in sequence along the direction away from the second area. When it is necessary to irrigate and fertilize the intercropped soybeans at the current time point, the historical light intensity information of all sub-areas in the first area is first obtained to provide data support for the subsequent prediction of light intensity. Then, considering the different degrees of influence of corn shading on different sub-areas and the differences in the fertilization cycles of soybeans at different growth stages, according to these historical light intensity information, the light intensity prediction information of each sub-area in the next fertilization cycle of soybeans after the current time point is determined using a pre-trained neural network model to achieve refined and differentiated dynamic prediction. Finally, according to the light intensity prediction information, the irrigation and fertilization scheme of each sub-area at the current time point is determined from a preset set of irrigation and fertilization schemes to achieve a personalized response to the growth needs of soybeans under different light conditions. Thus, the present application accurately considers the shading effect of corn on soybeans and achieves the purpose of refined irrigation and fertilization that is adapted to local conditions and time. This solves the technical problem that the irrigation and fertilization scheme in the corn-soybean intercropping mode in the prior art does not take into account the shading effect of corn on soybeans, which seriously affects the yield of intercropped soybeans.

[0029] Optionally, photosensitive sensors are arranged at intervals on each sub-area, and the intervals between the photosensitive sensors on all sub-areas arranged sequentially in a direction away from the second area increase sequentially; and the light intensity information of each sub-area is collected by the corresponding photosensitive sensors.

[0030] Specifically, since the sub-area close to corn (i.e., the part of the first area closer to the second area) is more significantly shaded, and the shadow effect of corn on soybeans changes with time and seasons, the change in light intensity is more complex and frequent. On the contrary, as the distance from the corn increases, the change tends to be gentle. In order to accurately reflect this change, the present application chooses to densely arrange photosensitive sensors in the sub-area close to the corn, and gradually relax the spacing between sensors in the sub-area far from the corn, referring to Figure 2a and Figure 2b As shown. This layout ensures that high-density data points are obtained in areas with the most variable lighting conditions, while avoiding excessive investment in hardware resources in areas with relatively stable lighting conditions. After that, the corresponding light sensors can continuously collect light intensity information for each sub-area in different time periods to provide a detailed record of light changes at different times and locations, providing solid data support for subsequent light intensity prediction and irrigation and fertilization strategy formulation.

[0031] In this way, even in an environment with rapidly changing light conditions, we can ensure that there are enough data points to depict the true distribution map of light intensity, which is crucial for subsequent light intensity prediction and the formulation of irrigation and fertilization plans. At the same time, this non-uniformly distributed sensor layout strategy effectively balances the accuracy and cost of data collection, avoiding the unnecessary economic burden of uniformly deploying a large number of sensors throughout the area.

[0032] Optionally, a drip irrigation pipe for irrigation and fertilization is installed at the location where the photosensitive sensor is provided in each sub-area.

[0033] Specifically, refer to Figure 2a and Figure 2b As shown in the figure, drip irrigation pipes for irrigation and fertilization are installed at the locations where photosensors are set in each sub-area. The drip irrigation pipes can achieve accurate delivery of water and fertilizers, avoiding the waste caused by traditional flood irrigation. In addition, by closely combining the monitoring point (photosensor) with the intervention point (drip irrigation pipe), the consistency and coherence between data collection and actual actions are ensured, so as to achieve independent control of each microenvironment and truly achieve refined management based on local conditions.

[0034] Optionally, based on the historical light intensity information, a pre-trained neural network model is used to determine the light intensity prediction information of each sub-area within a preset time after the current time point, including: determining the time length within the preset time after the current time point; for each sub-area, inputting the historical light intensity information of the sub-area and the time length into the pre-trained neural network model to determine the light intensity prediction information of the sub-area within the preset time after the current time point.

[0035] Specifically, considering the differences in the fertilization cycles of soybeans at different growth stages, for example, but not limited to, the fertilization cycle of soybean germination period is 7 days, the fertilization cycle of branching period is 10 days, the fertilization cycle of flowering and pod setting period is 12 days, and the fertilization cycle of maturity period is 15 days. Therefore, before using the pre-trained neural network model for prediction, it is necessary to first clarify the target time period for prediction (corresponding to the preset time after the current time point), that is, the length of time from the current time point to the end of the next fertilization cycle of soybeans. Then, for each sub-area, the historical light intensity information of the sub-area and the aforementioned determined time length are input into the pre-trained neural network model to generate the light intensity prediction value of each sub-area in the target time period, thereby realizing the transformation from historical light intensity data to future light condition prediction, providing an important basis for subsequent irrigation and fertilization decisions.

[0036] Optionally, the neural network model is a recurrent neural network model, a unidirectional long short-term memory network model or a gated recurrent unit model.

[0037] Specifically, considering that the historical light intensity information of each sub-region will change over time, it presents obvious sequence characteristics. The recurrent neural network (RNN) can process time-dependent sequence data through its cyclic connection structure, so it can capture the temporal correlation of historical light intensity information and provide a basis for prediction. The unidirectional long short-term memory network (LSTM) effectively solves the gradient vanishing / explosion problem of RNN when processing long sequences by introducing a special gating mechanism (input gate, forget gate, output gate), which enables it to remember long-term dependencies and is very suitable for prediction tasks that need to consider the impact of historical data on the current state. GRU is a simplified version of LSTM. By merging the input gate and the forget gate into a single update gate, the complexity of the model is reduced and the training speed is accelerated. In addition, GRU still retains the ability to process long sequence data, making it perform well in scenarios that require fast response and low computing resources. Therefore, the neural network model proposed in this application is a recurrent neural network model, a unidirectional long short-term memory network model, or a gated recurrent unit model.

[0038] Optionally, the neural network model is trained in the following manner: historical light intensity sample data of all sub-areas of the first area during the entire growth cycle of soybeans are obtained; a training set of the neural network model is determined based on the historical light intensity sample data; wherein each sample data in the training set is the light intensity data of each sub-area over a period of time, and the label of each sample data is the light intensity data of a soybean fertilization cycle after the period of time; based on the training set, the neural network model is trained until the neural network model meets a preset training cutoff condition.

[0039] Specifically, it is necessary to collect historical light intensity data for all sub-areas in the first area during the entire soybean growth cycle. These data can be obtained through continuous monitoring by previously installed light sensors to ensure that the entire process from sowing to harvesting is covered, including light conditions under different weather conditions and seasonal changes. Since the collected raw data often contains noise and missing values, preprocessing can be performed, including removing outliers, filling missing data, and normalizing values ​​to eliminate dimensional effects and improve the stability and efficiency of model training.

[0040] Then, a training set is constructed. Each sample data in the training set represents the light intensity sequence of a sub-area during a certain period of time, and its corresponding label is the light intensity data of a soybean fertilization cycle after this period of time. This design enables the model to learn the pattern of light intensity evolution over time and its indicative significance for the light conditions of future fertilization cycles.

[0041] Secondly, determine the time length of the sample data (i.e., the length of the input sequence) and the time point of the prediction target (i.e., the time corresponding to the label). The selection of these two parameters should be based on the understanding of soybean growth habits and the mechanism of light influence, as well as the results of preliminary experiments. After that, select a suitable neural network architecture (such as RNN, LSTM, GRU, etc.) and set the initial parameters, including the number of hidden layer nodes, activation function type, loss function, and optimizer. Set training cutoff conditions, such as the prediction error of the model on the validation set is lower than a certain threshold, or the performance on the validation set is no longer significantly improved after several consecutive iterations. In addition, in order to avoid infinite loops, a maximum number of iterations is usually set. Once this upper limit is reached, the training will terminate even if the performance indicators are not fully met.

[0042] Finally, the model is trained for multiple iterations using the constructed training set. In each iteration, the model attempts to predict future light intensities based on the input light intensity sequence, and adjusts the weights through the back-propagation algorithm to minimize the gap between the predicted value and the true label. During the training process, a portion of the data should be reserved as a validation set to evaluate the generalization ability of the model and prevent overfitting. Depending on the performance on the validation set, the model structure or training parameters, such as learning rate, batch size, etc., may need to be adjusted. After training is completed, save the model version with the best performance for deployment in actual agricultural production. As new light data accumulates over time, the model should be retrained or fine-tuned regularly to adapt to new patterns of environmental changes and crop growth to ensure the accuracy of predictions and the effectiveness of decisions.

[0043] Therefore, through the above carefully designed training process, the neural network model can learn to extract valuable information from the historical light intensity data, ensuring the prediction accuracy and reliability of the model, so as to accurately predict the light intensity information in the future time period.

[0044] Optionally, different fertilization schemes in the fertilization scheme set are associated with different light intensity intervals and corresponding soybean growth cycles; and, based on the light intensity prediction information, determining the fertilization scheme for each sub-area at the current time point from the preset fertilization scheme set, including: determining the soybean growth cycle at the current time point; determining the light intensity interval to which the light intensity prediction information of each sub-area belongs; and using the fertilization scheme in the preset fertilization scheme set that is associated with the light intensity interval to which the light intensity prediction information of each sub-area belongs and the soybean growth cycle at the current time point as the fertilization scheme for each sub-area at the current time point.

[0045] Specifically, the light intensity can be divided into several levels, such as low light, medium light, and high light, based on historical light data and the actual growth performance of soybeans. Each level corresponds to the typical growth characteristics and water and fertilizer demand patterns of soybeans under the light conditions. For each light intensity level, a corresponding irrigation and fertilization plan is developed in combination with the growth stage, water and fertilizer demand of soybeans. For example, during the flowering and podding period of soybeans and under high light conditions, the irrigation amount is moderately reduced to avoid excessive water inhibiting root respiration, while increasing the proportion of phosphorus and potassium fertilizers to promote the formation and filling of grains. During the flowering and podding period of soybeans and under low light conditions, attention should be paid to increasing the proportion of nitrogen fertilizers to stimulate the intrinsic metabolic activity of the plants. Afterwards, all customized irrigation and fertilization plans are integrated into a comprehensive set of plans to ensure coverage of the entire growth cycle from sowing to harvesting to meet the nutritional needs of soybeans at different growth stages and light conditions. At the same time, through field trials and data analysis, the details of each plan are continuously optimized to make it more in line with actual production needs.

[0046] Therefore, when determining the irrigation and fertilization scheme for each sub-area at the current time point, it is necessary to determine the current growth cycle of soybeans. That is, at any given time point, it is first necessary to determine which growth stage the soybeans are in, which can be achieved by observing the appearance characteristics of the plants, consulting growth records, or using sensor data. The light intensity prediction information of each sub-area is mapped to the corresponding light intensity interval. Since the preset irrigation and fertilization scheme set contains multiple schemes, each scheme is bound to a specific light intensity interval and soybean growth cycle, so through a lookup table or rule engine, the irrigation and fertilization scheme that matches the current light intensity prediction information and growth cycle can be quickly located.

[0047] For example, but not limited to, the light intensity prediction information of the first sub-area falls within the medium light interval, and soybeans are in the flowering and pod-setting period at this time. Therefore, an irrigation and fertilization plan designed for medium light conditions and the flowering and pod-setting period can be determined from the preset irrigation and fertilization plan set. This plan emphasizes the supplementation of elements such as nitrogen, phosphorus, and potassium to promote flower opening and pod formation. At the same time, the irrigation water volume is adjusted taking into account the light level to avoid excessive evaporation or water accumulation.

[0048] For another example, the neural network model predicts that the light intensity in the second sub-area will be in the high light range in the next 7 days, and according to the growth model, the soybeans in this sub-area are in the flowering and podding stage. Then, a scheme designed for the flowering and podding stage under high light conditions can be selected from the preset irrigation and fertilization scheme set. The scheme recommends moderately reducing the amount of irrigation to avoid excessive water inhibiting root respiration, and increasing the proportion of phosphorus and potassium fertilizers to promote the formation and filling of grains.

[0049] In this way, it can be ensured that the irrigation and fertilization plan for each sub-area can closely match its actual light conditions and soybean growth needs, thereby achieving precise allocation of resources and significantly improving the efficiency and sustainability of agricultural production.

[0050] In addition, according to this embodiment, a storage medium is also provided, wherein the storage medium includes a stored program, wherein when the program is run, a processor executes any one of the above methods.

[0051] The present application takes into account that there is a clear difference in the shading effect between the position near the corn and the position far from the corn in the first area, and the first area is divided into at least two sub-areas arranged in sequence along the direction away from the second area. When it is necessary to irrigate and fertilize the intercropped soybeans at the current time point, the historical light intensity information of all sub-areas in the first area is first obtained to provide data support for the subsequent prediction of light intensity. Then, considering the different degrees of influence of corn shading on different sub-areas and the differences in the fertilization cycles of soybeans at different growth stages, according to these historical light intensity information, the light intensity prediction information of each sub-area in the next fertilization cycle of soybeans after the current time point is determined using a pre-trained neural network model to achieve refined and differentiated dynamic prediction. Finally, according to the light intensity prediction information, the irrigation and fertilization scheme of each sub-area at the current time point is determined from a preset set of irrigation and fertilization schemes to achieve a personalized response to the growth needs of soybeans under different light conditions. Thus, the present application accurately considers the shading effect of corn on soybeans and achieves the purpose of refined irrigation and fertilization that is adapted to local conditions and time. This solves the technical problem that the irrigation and fertilization scheme in the corn-soybean intercropping mode in the prior art does not take into account the shading effect of corn on soybeans, which seriously affects the yield of intercropped soybeans.

[0052] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0053] Example 2 Figure 4 The schematic diagram of the structure of the device 400 for optimizing the intercropping of corn and soybean according to the present embodiment is shown. The device 400 includes: an information acquisition module 410, which is used to acquire the historical light intensity information of the first area at the current time point; wherein the first area is planted with soybeans, and the second area adjacent to the first area is planted with corn; the first area includes at least two sub-areas arranged in sequence in a direction away from the second area; the historical light intensity information includes the historical light intensity information of all sub-areas; an information determination module 420, which is used to determine the light intensity prediction information of each sub-area within a preset time after the current time point using a pre-trained neural network model according to the historical light intensity information; wherein the preset time is the next fertilization cycle of the soybean; and a scheme determination module 430, which is used to determine the irrigation and fertilization scheme of each sub-area at the current time point from a preset set of irrigation and fertilization schemes according to the light intensity prediction information.

[0054] Optionally, photosensitive sensors are arranged at intervals on each sub-area, and the intervals between the photosensitive sensors on all sub-areas arranged sequentially in a direction away from the second area increase sequentially; and the light intensity information of each sub-area is collected by the corresponding photosensitive sensors.

[0055] Optionally, a drip irrigation pipe for irrigation and fertilization is installed at the location where the photosensitive sensor is provided in each sub-area.

[0056] Optionally, based on the historical light intensity information, a pre-trained neural network model is used to determine the light intensity prediction information of each sub-area within a preset time after the current time point, including: determining the time length within the preset time after the current time point; for each sub-area, inputting the historical light intensity information of the sub-area and the time length into the pre-trained neural network model to determine the light intensity prediction information of the sub-area within the preset time after the current time point.

[0057] Optionally, the neural network model is a recurrent neural network model, a unidirectional long short-term memory network model or a gated recurrent unit model.

[0058] Optionally, the neural network model is trained in the following manner: historical light intensity sample data of all sub-areas of the first area during the entire growth cycle of soybeans are obtained; a training set of the neural network model is determined based on the historical light intensity sample data; wherein each sample data in the training set is the light intensity data of each sub-area over a period of time, and the label of each sample data is the light intensity data of a soybean fertilization cycle after the period of time; based on the training set, the neural network model is trained until the neural network model meets a preset training cutoff condition.

[0059] Optionally, different fertilization schemes in the fertilization scheme set are associated with different light intensity intervals and corresponding soybean growth cycles; and, based on the light intensity prediction information, determining the fertilization scheme for each sub-area at the current time point from the preset fertilization scheme set, including: determining the soybean growth cycle at the current time point; determining the light intensity interval to which the light intensity prediction information of each sub-area belongs; and using the fertilization scheme in the preset fertilization scheme set that is associated with the light intensity interval to which the light intensity prediction information of each sub-area belongs and the soybean growth cycle at the current time point as the fertilization scheme for each sub-area at the current time point.

[0060] Thus, according to this embodiment, considering that there is a clear difference between the shading effect of the position near the corn and the position far from the corn in the first area, the first area is divided into at least two sub-areas arranged in sequence along the direction away from the second area. When it is necessary to irrigate and fertilize the intercropped soybean at the current time point, the historical light intensity information of all sub-areas in the first area is first obtained to provide data support for the subsequent prediction of light intensity. Then, considering the different degrees of corn shading in different sub-areas and the differences in the fertilization cycle of soybeans at different growth stages, according to these historical light intensity information, the light intensity prediction information of each sub-area in the next fertilization cycle of soybeans after the current time point is determined using a pre-trained neural network model to achieve refined and differentiated dynamic prediction. Finally, according to the light intensity prediction information, the irrigation and fertilization scheme of each sub-area at the current time point is determined from a preset set of irrigation and fertilization schemes to achieve a personalized response to the growth needs of soybeans under different light conditions. Thus, the present application accurately considers the shading effect of corn on soybeans and achieves the purpose of refined irrigation and fertilization that is adapted to local conditions and time. This solves the technical problem that the irrigation and fertilization scheme in the corn-soybean intercropping mode in the prior art does not take into account the shading effect of corn on soybeans, which seriously affects the yield of intercropped soybeans.

[0061] Example 3 Figure 5 The irrigation and fertilization optimization system 500 for intercropping corn and soybeans according to the present embodiment is shown, comprising: a processor 510; and a memory 520, connected to the processor 510, for providing the processor 510 with instructions for processing the following processing steps: at a current time point, obtaining historical light intensity information of a first area; wherein the first area is planted with soybeans, and a second area adjacent to the first area is planted with corn; the first area includes at least two sub-areas arranged in sequence in a direction away from the second area; the historical light intensity information includes historical light intensity information of all sub-areas; based on the historical light intensity information, determining light intensity prediction information of each sub-area within a preset time after the current time point using a pre-trained neural network model; wherein the preset time is the next fertilization cycle of the soybeans; and based on the light intensity prediction information, determining an irrigation and fertilization scheme for each sub-area at the current time point from a preset set of irrigation and fertilization schemes.

[0062] Optionally, photosensitive sensors are arranged at intervals on each sub-area, and the intervals between the photosensitive sensors on all sub-areas arranged sequentially in a direction away from the second area increase sequentially; and the light intensity information of each sub-area is collected by the corresponding photosensitive sensors.

[0063] Optionally, a drip irrigation pipe for irrigation and fertilization is installed at the location where the photosensitive sensor is provided in each sub-area.

[0064] Optionally, based on the historical light intensity information, a pre-trained neural network model is used to determine the light intensity prediction information of each sub-area within a preset time after the current time point, including: determining the time length within the preset time after the current time point; for each sub-area, inputting the historical light intensity information of the sub-area and the time length into the pre-trained neural network model to determine the light intensity prediction information of the sub-area within the preset time after the current time point.

[0065] Optionally, the neural network model is a recurrent neural network model, a unidirectional long short-term memory network model or a gated recurrent unit model.

[0066] Optionally, the neural network model is trained in the following manner: historical light intensity sample data of all sub-areas of the first area during the entire growth cycle of soybeans are obtained; a training set of the neural network model is determined based on the historical light intensity sample data; wherein each sample data in the training set is the light intensity data of each sub-area over a period of time, and the label of each sample data is the light intensity data of a soybean fertilization cycle after the period of time; based on the training set, the neural network model is trained until the neural network model meets a preset training cutoff condition.

[0067] Optionally, different fertilization schemes in the fertilization scheme set are associated with different light intensity intervals and corresponding soybean growth cycles; and, based on the light intensity prediction information, determining the fertilization scheme for each sub-area at the current time point from the preset fertilization scheme set, including: determining the soybean growth cycle at the current time point; determining the light intensity interval to which the light intensity prediction information of each sub-area belongs; and using the fertilization scheme in the preset fertilization scheme set that is associated with the light intensity interval to which the light intensity prediction information of each sub-area belongs and the soybean growth cycle at the current time point as the fertilization scheme for each sub-area at the current time point.

[0068] Thus, according to this embodiment, considering that there is a clear difference between the shading effect of the position near the corn and the position far from the corn in the first area, the first area is divided into at least two sub-areas arranged in sequence along the direction away from the second area. When it is necessary to irrigate and fertilize the intercropped soybean at the current time point, the historical light intensity information of all sub-areas in the first area is first obtained to provide data support for the subsequent prediction of light intensity. Then, considering the different degrees of corn shading in different sub-areas and the differences in the fertilization cycle of soybeans at different growth stages, according to these historical light intensity information, the light intensity prediction information of each sub-area in the next fertilization cycle of soybeans after the current time point is determined using a pre-trained neural network model to achieve refined and differentiated dynamic prediction. Finally, according to the light intensity prediction information, the irrigation and fertilization scheme of each sub-area at the current time point is determined from a preset set of irrigation and fertilization schemes to achieve a personalized response to the growth needs of soybeans under different light conditions. Thus, the present application accurately considers the shading effect of corn on soybeans and achieves the purpose of refined irrigation and fertilization that is adapted to local conditions and time. This solves the technical problem that the irrigation and fertilization scheme in the corn-soybean intercropping mode in the prior art does not take into account the shading effect of corn on soybeans, which seriously affects the yield of intercropped soybeans.

[0069] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0070] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0071] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0072] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0073] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0074] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk, etc. Various media that can store program codes.

[0075] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for optimizing irrigation and fertilization for intercropping of corn and soybean, characterized in that: include: At the current time point, obtain historical light intensity information of the first area; The first area is planted with soybeans, and the second area adjacent to the first area is planted with corn; the first area includes at least two sub-areas arranged in sequence in a direction away from the second area; the historical light intensity information includes the historical light intensity information of all sub-areas; According to the historical light intensity information, a pre-trained neural network model is used to determine the light intensity prediction information of each sub-area within a preset time after the current time point; wherein the preset time is the next fertilization cycle of the soybean; as well as According to the light intensity prediction information, the fertigation scheme for each sub-area at the current time point is determined from a preset fertigation scheme set.

2. The method according to claim 1, characterized in that: Photosensitive sensors are arranged at intervals on each sub-area, and the intervals between the photosensitive sensors on all sub-areas arranged sequentially in a direction away from the second area increase sequentially; the light intensity information of each sub-area is collected by the corresponding photosensitive sensors.

3. The method according to claim 2, characterized in that A drip irrigation pipe for irrigation and fertilization is installed at the location where the photosensitive sensor is set in each sub-area.

4. The method according to claim 1, characterized in that The method of determining the light intensity prediction information of each sub-area within a preset time after the current time point using a pre-trained neural network model according to the historical light intensity information includes: Determine the length of time within a preset time after the current time point; For each sub-region, the historical light intensity information of the sub-region and the time length are input into the pre-trained neural network model to determine the light intensity prediction information of the sub-region within a preset time after the current time point.

5. The method according to claim 4, characterized in that The neural network model is a recurrent neural network model, a unidirectional long short-term memory network model or a gated recurrent unit model.

6. The method according to claim 4, characterized in that The neural network model is trained in the following way: Obtaining historical light intensity sample data of all sub-areas of the first area during the entire growth period of soybeans; Determine a training set of the neural network model according to the historical light intensity sample data; wherein each sample data in the training set is light intensity data of each sub-region within a period of time, and the label of each sample data is light intensity data of a soybean fertilization cycle after the period of time; Based on the training set, the neural network model is trained until the neural network model meets a preset training cutoff condition.

7. The method according to claim 1, characterized in that Different fertilization schemes in the fertilization scheme set are associated with different light intensity intervals and corresponding soybean growth cycles; and, determining the fertilization scheme for each sub-area at the current time point from the preset fertilization scheme set according to the light intensity prediction information, includes: Determining the growth cycle of soybeans at the current time point; Determine the illumination intensity interval to which the illumination intensity prediction information of each sub-region belongs; The fertilization scheme in the preset fertilization scheme set that is associated with the light intensity interval to which the light intensity prediction information of each sub-region belongs and the growth cycle of soybean at the current time point is used as the fertilization scheme for each sub-region at the current time point.

8. A storage medium, characterized in that: The storage medium includes a stored program, wherein when the program is run, the processor executes the method according to any one of claims 1 to 7.

9. An irrigation and fertilization optimization device for corn and soybean intercropping, comprising: An information acquisition module, used to acquire historical light intensity information of the first area at a current time point; The first area is planted with soybeans, and the second area adjacent to the first area is planted with corn; the first area includes at least two sub-areas arranged in sequence in a direction away from the second area; the historical light intensity information includes the historical light intensity information of all sub-areas; An information determination module, used to determine the light intensity prediction information of each sub-area within a preset time after the current time point based on the historical light intensity information using a pre-trained neural network model; wherein the preset time is the next fertilization cycle of the soybean; as well as The scheme determination module is used to determine the irrigation and fertilization scheme for each sub-area at the current time point from a preset irrigation and fertilization scheme set according to the light intensity prediction information.

10. An irrigation and fertilization optimization system for corn and soybean intercropping, characterized in that: include: processor; as well as A memory, connected to the processor, configured to provide the processor with instructions for processing the following processing steps: At the current time point, historical light intensity information of a first area is obtained; wherein soybeans are planted in the first area, and corn is planted in a second area adjacent to the first area; the first area includes at least two sub-areas arranged in sequence in a direction away from the second area; and the historical light intensity information includes historical light intensity information of all sub-areas; According to the historical light intensity information, a pre-trained neural network model is used to determine the light intensity prediction information of each sub-area within a preset time after the current time point; wherein the preset time is the next fertilization cycle of the soybean; as well as According to the light intensity prediction information, the fertigation scheme for each sub-area at the current time point is determined from a preset fertigation scheme set.

Images