An evaluation method and system for the improvement effect of erosion resistance and soil fertility improvement in degenerated black soil areas

Abstract

The invention relates to the technical field of evaluation of cultivated land improvement effects, and specifically discloses a method and system for evaluating the anti-erosion fertilization improvement effect in a degraded black soil area. The method comprises receiving an improvement behavior input by a user, and determining an evaluation level of the improvement behavior according to historical behavior; selecting sampling points in a black soil improvement area, and acquiring soil data containing time tags at the sampling points; identifying the soil data containing time tags, generating improvement effects, and synchronously adjusting the data acquisition frequency of the sampling points; acquiring terrain parameters at each sampling point, constructing a sample set of terrain parameters and improvement effects under the improvement behavior, and training a neural network model; the invention selects sampling points in a degraded black soil area, and synchronously introduces a dynamic data sampling frequency, so that under the same cost condition, more data of important sampling points are acquired, the data quantity and quality are very high, and the cost is reduced in disguise. When the cost is limited, more test areas can be built.

Description

Technical Field

[0001] The present invention relates to the technical field of evaluating the effect of cultivated land improvement, and particularly to a method and system for evaluating the effect of erosion resistance and fertility improvement in degraded black soil areas. Background Art

[0002] The erosion resistance and fertility improvement in degraded black soil areas is an important agricultural and ecological issue, mainly involving enhancing the soil's erosion resistance ability, restoring soil fertility, and improving the sustainability of farmland ecosystems.

[0003] There are many ways to improve degraded black soil areas, and with the progress of technology, the improvement methods are constantly innovating. For each improvement method, it is necessary to first evaluate its effect and then determine whether to apply it on a large scale. Since the improvement process is affected by a large number of parameters, including various environmental parameters, the effect evaluation process is very difficult and often requires actual evaluation in the test area, with high costs. If the costs can be reduced, then under the same cost conditions, more test areas can be set up to improve the accuracy of the evaluation results. How to reduce the costs of the evaluation process is the technical problem that the technical solution of the present invention aims to solve. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for evaluating the effect of erosion resistance and fertility improvement in degraded black soil areas to solve the problems raised in the above background art.

[0005] To achieve the above purpose, the present invention provides the following technical solutions:

[0006] A method for evaluating the effect of erosion resistance and fertility improvement in degraded black soil areas, the method comprising:

[0007] Receiving the improvement behavior input by the user and determining the evaluation level of the improvement behavior according to the historical behavior;

[0008] Selecting sampling points in the black soil improvement area and obtaining soil data with time tags at the sampling points;

[0009] Identifying the soil data with time tags to generate the improvement effect and synchronously adjusting the data acquisition frequency at the sampling points; the data acquisition frequency is proportional to the data weight at the sampling points, and the data weight is used to characterize the difference degree between the data and other data; the minimum value of the data acquisition frequency is determined by the evaluation level;

[0010] Obtaining the terrain parameters at each sampling point, constructing a sample set of the terrain parameters and the improvement effect under the improvement behavior, training a neural network model, and when the error rate of the neural network model is less than the preset error rate threshold, outputting the trained neural network model as the evaluation model.

[0011] As a further solution of the present invention, the steps of receiving the improved behavior input by the user and determining the evaluation level of the improved behavior according to the historical behavior include:

[0012] Receiving the improved behavior input by the user; the improved behavior is in the form of an array, and the serial number of each element in the array represents the order of the operation executed, and the value of the element is a number; the number contains a sub-element representing the operation type and a sub-element representing the operation amount;

[0013] Counting each improved behavior in each black soil improvement area to obtain historical behavior; among them, the improved behavior in the historical behavior contains a time tag;

[0014] Comparing the improved behavior input by the user with the historical behavior to determine the evaluation level of the improved behavior;

[0015] The process of determining the evaluation level is:

[0016] In the formula, is the evaluation level, is a preset gradient function, represents the independent variable, the threshold in the gradient function is a preset value, and represents the number of behaviors in the historical behavior that reach the preset similarity threshold with the improved behavior

[0017] As a further solution of the present invention, the steps of selecting sampling points in the black soil improvement area and obtaining soil data with time tags at the sampling points include:

[0018] Querying the crop planting positions in the black soil improvement area and determining the sampling area based on the crop planting positions;

[0019] Randomly selecting a preset number of points in the sampling area as sampling points;

[0020] Installing a data collector at the sampling point to obtain soil data. The soil data at least includes the visual information of the crop and the physical and chemical data of the soil

[0021] As a further solution of the present invention, the steps of identifying the soil data with time tags, generating an improvement effect, and synchronously adjusting the data acquisition frequency of the sampling points include:

[0022] For any type of soil data, sorting it according to the time tag to obtain a data sequence;

[0023] Regularizing the data sequence based on preset standard data to obtain a regularized sequence;

[0024] Obtaining the difference sequence of the regularized sequence to generate an improvement effect; the improvement effect is a parameter jointly determined by the regularized sequence and the difference sequence;

[0025] Compare all regularized sequences of different sampling points, and synchronously adjust the data acquisition frequency of the sampling points according to the comparison results; wherein, the comparison process of the regularized sequences occurs between soil data of the same type.

[0026] As a further solution of the present invention: the steps of comparing all regularized sequences of different sampling points and synchronously adjusting the data acquisition frequency of the sampling points according to the comparison results include:

[0027] For any sampling point, sequentially determine the type of soil data;

[0028] Read the regularized sequence of this sampling point and the regularized sequences of other sampling points under this type, compare the regularized sequences, and calculate the similarity and its similarity mean value;

[0029] Statistically calculate the similarity mean value corresponding to each type, and determine the data acquisition frequency of the sampling point;

[0030] The process of determining the data acquisition frequency is:

[0031] ; where represents the data acquisition frequency, is a preset constant, is the total number of types, is the preset correction coefficient of the th type, is the th type of similarity mean value.

[0032] As a further solution of the present invention: the steps of obtaining the terrain parameters at each sampling point, constructing a sample set of terrain parameters and improvement effects under the improvement behavior, training a neural network model, and when the error rate of the neural network model is less than the preset error rate threshold, outputting the trained neural network model as an evaluation model include:

[0033] Obtain the remote sensing data of the sampling area corresponding to each sampling point based on the remote sensing data;

[0034] Identify the remote sensing data, extract the terrain contour map as the terrain parameter;

[0035] Read the improvement effect of the sampling point, and establish a sample set from the terrain parameter to the improvement effect;

[0036] Split the sample set into a training set, a test set and a validation set, and train the neural network model;

[0037] When the error rate of the neural network model is less than the preset error rate threshold, output the trained neural network model as an evaluation model.

[0038] The technical solution of the present invention also provides an evaluation system for the anti-erosion and fertility improvement effect in degraded black soil areas, and the system includes:

[0039] An evaluation level determination module, configured to receive the improvement behavior input by the user, and determine the evaluation level of the improvement behavior according to the historical behavior;

[0040] A soil data acquisition module, configured to select sampling points in the black soil improvement area, and acquire soil data with time tags at the sampling points;

[0041] A data application module, configured to identify the soil data with time tags, generate an improvement effect, and synchronously adjust the data acquisition frequency of the sampling points; the data acquisition frequency is proportional to the data weight at the sampling point, and the data weight is used to characterize the difference degree between the data and other data; the minimum value of the data acquisition frequency is determined by the evaluation level;

[0042] A model training module, configured to acquire the terrain parameters at each sampling point, construct a sample set of the terrain parameters and the improvement effect under the improvement behavior, train a neural network model, and when the error rate of the neural network model is less than a preset error rate threshold, output the trained neural network model as an evaluation model.

[0043] As a further solution of the present invention: the evaluation level determination module includes:

[0044] A behavior recording unit, configured to receive the improvement behavior input by the user; the improvement behavior is in the form of an array, and the serial number of each element in the array represents the order of the executed operation, and the value of the element is a number; the number contains a sub-element representing the operation type and a sub-element representing the operation amount;

[0045] A behavior statistics unit, configured to count each improvement behavior in each black soil improvement area to obtain historical behavior; wherein, the improvement behavior in the historical behavior contains a time tag;

[0046] A comparison and application unit, configured to compare the improvement behavior input by the user with the historical behavior to determine the evaluation level of the improvement behavior;

[0047] The process of determining the evaluation level is:

[0048] ; where is the evaluation level, is a preset gradient function, represents the independent variable, and the threshold in the gradient function is a preset value, represents the number of behaviors in the historical behavior whose similarity to the improvement behavior reaches a preset similarity threshold.

[0049] As a further solution of the present invention: the soil data acquisition module includes:

[0050] A sampling area determination unit, used to query the crop planting location in the black soil improvement area and determine the sampling area based on the crop planting location;

[0051] A sampling point selection unit is used to randomly select a preset number of points in the sampling area as sampling points;

[0052] The acquisition unit is used to install a data collector at a sampling point to acquire soil data; the soil data at least includes visual information of the crop and physical and chemical data of the soil.

[0053] As a further solution of the present invention: the data application module includes:

[0054] A data sorting unit is used to sort any type of soil data according to the time label to obtain a data sequence;

[0055] A data regularization unit, used to regularize the data sequence based on preset standard data to obtain a regularized sequence;

[0056] A difference calculation unit, used to obtain a difference sequence of the regular sequence and generate an improved effect; the improved effect is a parameter determined by the regular sequence and the difference sequence;

[0057] The frequency adjustment unit is used to compare all regular sequences of different sampling points and synchronously adjust the data acquisition frequency of the sampling points according to the comparison results; wherein the comparison process of the regular sequences occurs between soil data of the same type.

[0058] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention selects sampling points in the degraded black soil area and simultaneously introduces a dynamic data sampling frequency, so that under the same cost conditions, more data from important sampling points can be obtained, and the data quantity and quality are very high, which reduces the cost in disguise. When the cost is limited, more test areas can be built. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention.

[0060] Figure 1 This is a flowchart of the evaluation method for the effect of anti-erosion fertilization improvement in degraded black soil areas.

[0061] Figure 2 This is a structural diagram of the system for evaluating the effect of anti-erosion fertilization improvement in degraded black soil areas. DETAILED DESCRIPTION

[0062] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention more clear and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0063] Figure 1 FIG. is a flowchart of an evaluation method for the improvement effect of anti-erosion and fertility improvement in degraded black soil areas. In an embodiment of the present invention, an evaluation method for the improvement effect of anti-erosion and fertility improvement in degraded black soil areas, the method includes:

[0064] Step S100: Receive the improvement behavior input by the user, and determine the evaluation level of the improvement behavior according to historical behaviors;

[0065] The improvement behavior is an act of improving the degraded black soil area. There are many ways of improvement, including planting behaviors such as straw returning to the field, contour planting, crop rotation and intercropping, fertilization behaviors such as applying organic fertilizers, planting green manure crops and applying biological fertilizers, and modification behaviors such as applying lime or gypsum. These behaviors are collectively referred to as improvement behaviors; every time the management party makes an improvement, there will be an improvement behavior. Subsequently, soil data is obtained and analyzed to evaluate the quality of the improvement behavior. In this application, it is to obtain the improvement effect; among them, all known improvement behaviors are counted as historical behaviors. When receiving the improvement behavior input by the user, it is compared with the historical behaviors to determine whether the current improvement behavior is rare, and then the evaluation level is determined. The rarer the improvement behavior, the higher the evaluation level (the higher it means that more resources need to be invested in evaluation).

[0066] Step S200: Select sampling points in the black soil improvement area, and obtain soil data with time tags at the sampling points;

[0067] Select sampling points in the black soil improvement area, and install data acquisition equipment at the sampling points to obtain soil data. When obtaining soil data, the time of acquisition needs to be recorded as a time tag; among them, the data acquisition equipment installed at the sampling points can use existing data acquisition equipment, and the obtained data is also related to the equipment.

[0068] Step S300: Identify the soil data with time tags to generate an improvement effect, and synchronously adjust the data acquisition frequency of the sampling points; the data acquisition frequency is proportional to the data weight at the sampling points, and the data weight is used to characterize the degree of difference between the data and other data; the minimum value of the data acquisition frequency is determined by the evaluation level;

[0069] Identify soil data with time tags. On the one hand, evaluate the soil data itself. On the other hand, also evaluate the change of soil data (since the time tags are known, just calculate the change amount). Combining the two evaluation results, the improvement effect at the sampling point where the soil data is located can be determined. Combining the improvement effects at all sampling points, the final improvement effect can be obtained. The improvement effect can also be called the improvement effect.

[0070] Meanwhile, each identification process occurs at each sampling point. According to its evaluation process, the importance of each sampling point in the evaluation process can be determined, that is, which sampling points have a more important evaluation process. Thus, update the data acquisition frequency so that more important sampling points can obtain more data. This architecture can obtain more useful data under the same cost conditions.

[0071] Specifically, regarding the importance of each sampling point, the importance can be evaluated by its difference from other sampling points. The greater the difference between a sampling point and other sampling points, the higher the importance.

[0072] Step S400: Obtain the terrain parameters at each sampling point, construct a sample set of terrain parameters and improvement effects under the improvement behavior, train a neural network model. When the error rate of the neural network model is less than the preset error rate threshold, output the trained neural network model as an evaluation model.

[0073] Obtain the terrain parameters at each sampling point. The data structure of the terrain parameters can be determined by the user in advance. It can be some parameter sets, such as height, maximum height difference, and soil type, etc. It can also be some other formats, such as the three-dimensional model of the soil. Existing AI models have extremely strong data processing capabilities. No matter what type of data it is, the AI model can recognize it. This also means that with the help of AI, a local simplified model that can be oriented to multiple formats can be trained.

[0074] After obtaining the terrain parameters at each sampling point in this application, read the corresponding improvement effect, establish a sample set from terrain parameters to improvement effects, and train a neural network model. In this process, AI can be introduced to transcode the terrain parameters, etc. When the error rate of the trained neural network model is less than the preset error rate threshold, output the trained neural network model as an evaluation model. It should be noted that the obtained evaluation model is the evaluation model corresponding to each improvement behavior. After training, for the improvement behavior input by the user, an improvement effect can be predicted based on the actual terrain parameters. This prediction is instant and has extremely high accuracy. Compared with the actual long-term evaluation process, the accuracy is slightly insufficient, but the efficiency is greatly improved (from several months to within one second).

[0075] Regarding step S100, the steps of receiving the improved behavior input by the user and determining the evaluation level of the improved behavior according to historical behaviors include:

[0076] Receiving the improved behavior input by the user; the improved behavior is in the form of an array, and the serial number of each element in the array represents which operation is executed, and the value of the element is a number; the number contains a sub-element representing the operation type and a sub-element representing the operation quantity;

[0077] Counting each improved behavior in each black soil improvement area to obtain historical behaviors; among them, the improved behaviors in the historical behaviors contain time tags;

[0078] Comparing the improved behavior input by the user with the historical behaviors to determine the evaluation level of the improved behavior.

[0079] Receiving the improved behavior input by the user, as can be seen from the foregoing, the improved behavior is actually a kind of behavior, which includes multiple behaviors and is not a simple number. In the computer processing process, it is not very convenient. Therefore, this application digitizes the improved behavior. First, each improved behavior is numbered, and then the specific quantity of improvement resources is determined, so as to obtain the improved behavior in digital form; for example, assuming that the improved behavior includes 8 digits, the first 4 digits or the first 2 digits represent the number corresponding to the improved behavior, and the last 4 digits or the last 6 digits represent the quantity of improvement resources of the improved behavior. If the improved behavior is fertilization, the quantity of improvement resources is the quantity of fertilizer.

[0080] Counting each improved behavior in each black soil improvement area to obtain historical behaviors. When counting the improved behaviors, record the improvement duration, generate time tags, and when receiving the improved behavior input by the user, compare the improved behavior input by the user with the historical behaviors to determine the evaluation level of the improved behavior.

[0081] In an example of the technical solution of the present invention, the process of determining the evaluation level is:

[0082] ; where is the evaluation level, is a preset gradient function, represents the independent variable, and the threshold value in the gradient function is a preset value, represents the number of behaviors in the historical behaviors whose similarity to the improved behavior reaches the preset similarity threshold.

[0083] The evaluation process is very simple. First, obtain the number of behaviors in the historical behaviors that are similar enough to the improvement behavior. Based on the quantity threshold reached by the number, the evaluation level can be determined. Specifically, regarding how to compare behaviors, after the improvement behavior is numbered as above, perform a logical AND operation on the first 4 digits or the first 2 digits to determine behaviors of the same type, and then perform a difference operation on the last 4 digits or the last 6 digits. The similarity is determined by the inverse ratio of the difference, indicating whether the improvement resource amounts are similar.

[0084] Regarding step S200, the step of selecting sampling points in the black soil improvement area and obtaining soil data with time tags at the sampling points includes:

[0085] Query the crop planting positions in the black soil improvement area, and determine the sampling area based on the crop planting positions;

[0086] Randomly select a preset number of points in the sampling area as sampling points;

[0087] Install a data collector at the sampling point to obtain soil data; the soil data includes at least the visual information of the crop and the physical and chemical data of the soil.

[0088] The evaluation process of the black soil improvement area depends on the growth process of plants. The improvement effect of an area will ultimately be reflected in the crops. The growth trend of the crops indicates the soil quality. Based on this principle, query the crop planting positions in the black soil improvement area, determine the sampling area based on the crop planting positions. The sampling area can be a circular area centered on the crop planting positions. Randomly select a preset number of points in the sampling area as sampling points. This is equivalent to selecting some points with the crop planting positions as the core in the black soil improvement area. Install a data collector at the sampling point to obtain soil data. In the technical solution of the present invention, a type of soil data is defined, which is the visual information of the crop. Correspondingly, the data sampler also includes a camera. As for other soil data, select a part from the existing physical and chemical data of the soil. Correspondingly, the data collector also needs to use the corresponding equipment.

[0089] Regarding step S300, the step of identifying the soil data with time tags to generate an improvement effect and synchronously adjusting the data acquisition frequency of the sampling points includes:

[0090] For any type of soil data, sort it according to the time tag to obtain a data sequence;

[0091] Regularize the data sequence based on the preset standard data to obtain a regularized sequence;

[0092] Obtain the difference sequence of the regularized sequence to generate an improvement effect; the improvement effect is a parameter jointly determined by the regularized sequence and the difference sequence;

[0093] Compare all the regularized sequences of different sampling points, and synchronously adjust the data acquisition frequency of the sampling points according to the comparison results; among them, the comparison process of the regularized sequences occurs between soil data of the same type.

[0094] In an example of the technical solution of the present invention, the data analysis process is defined. For any type of soil data, it is sorted according to the time tags to obtain a data sequence, and the data sequence is regularized based on a preset standard data to obtain a regularized sequence. The regularization process can use the ratio, that is, each element in the data sequence is divided by the standard data. The calculated regularized data itself is also a data without units. This also means that for different types of soil data, regularized sequences of the same format can be obtained.

[0095] Perform further processing on the regularized sequence, calculate the difference to obtain a difference sequence. At this time, the regularized sequence reflects the actual situation, and the difference sequence reflects the change situation. Combine the regularized sequence and the difference sequence to generate a parameter called the improvement effect.

[0096] It is worth mentioning that the soil data in this application also includes the visual information of the crops. In practical applications, a visual recognition algorithm also needs to be introduced to extract the growth parameters of the crops, such as chlorophyll content, etc. At this time, the visual information is also converted into a data sequence, and then the same calculation process is used for calculation.

[0097] As a preferred embodiment of the technical solution of the present invention, this application will also determine the data acquisition frequency of the sampling points according to the regularized sequence. The determination process involves the comparison process of the regularized sequence, but the comparison process is a comparison between the same types, that is, only two regularized sequences of the same type of soil data will be compared, otherwise the comparison itself is meaningless. Specifically, the step of comparing all the regularized sequences of different sampling points and synchronously adjusting the data acquisition frequency of the sampling points according to the comparison results includes:

[0098] For any sampling point, sequentially determine the type of soil data;

[0099] Read the regularized sequence of this sampling point and the regularized sequences of other sampling points under this type, compare the regularized sequences, and calculate the similarity and its similarity mean value;

[0100] Statistically analyze the similarity mean values corresponding to each type to determine the data acquisition frequency of the sampling points.

[0101] For any sampling point, there will be multiple types of soil data. Process the multiple types of soil data sequentially. For each type, read the regular sequence of this sampling point under this type, and then read the corresponding regular sequences of other sampling points of the same type. By comparing the regular sequences, the similarity between the selected sampling point and other sampling points under this type can be obtained. Since there are a large number of other sampling points, multiple similarities can be obtained. Calculate the mean value of the similarities to obtain the similarity mean value. The similarity reflects the similarity situation between this sampling point and other sampling points under this type.

[0102] Statistically calculate the mean values of similarities corresponding to all types of soil data at the sampling point, and then the comprehensive similarity situation between the sampling point and other sampling points can be obtained, thereby determining the data acquisition frequency.

[0103] Regarding the data acquisition frequency, the process of determining the data acquisition frequency is as follows:

[0104] ; where represents the data acquisition frequency, is a preset constant, is the total number of types, is the correction coefficient of the th type preset, is the th mean value of similarity.

[0105] After calculating the mean values of similarities, sum up all the mean values of similarities to obtain the total mean value of similarities. Determine the data acquisition frequency according to the inverse ratio of the total mean value of similarities. That is, the larger the mean value of similarities, the more common this sampling point is, and the smaller the data acquisition frequency. On this basis, the present application also includes a correction coefficient term, which represents the importance degree of different types of soil data and is preset by the user.

[0106] Regarding step S400, the steps of obtaining the terrain parameters at each sampling point, constructing a sample set of terrain parameters and improvement effects under the improvement behavior, and training a neural network model, and when the error rate of the neural network model is less than the preset error rate threshold, outputting the trained neural network model as an evaluation model include:

[0107] Obtain the remote sensing data of the sampling area corresponding to each sampling point based on remote sensing data;

[0108] Identify the remote sensing data, extract the terrain contour map as the terrain parameter;

[0109] Read the improvement effect of the sampling point and establish a sample set from terrain parameters to improvement effects;

[0110] Divide the sample set into a training set, a test set and a validation set, and train the neural network model;

[0111] When the error rate of the neural network model is less than a preset error rate threshold, the trained neural network model is output as an evaluation model.

[0112] The training process of the neural network model is not complicated. Using the terrain parameters as features and the improvement effect as labels, a sample is constructed. When the number of samples is large enough, a sample set is obtained. The sample set is divided into a training set, a test set, and a validation set to train the neural network model. When the error rate of the neural network model is less than a preset error rate threshold, the trained neural network model is output as an evaluation model.

[0113] In the above content, the terrain parameters are defined. In fact, the terrain parameters themselves are a higher-level concept. Those skilled in the art can use the average height as the terrain parameter, which is specifically determined by the user, and the present application will not elaborate. In the above content, the terrain parameters are defined as the terrain contour map extracted from remote sensing data. The terrain contour map is the contour recognition result of the remote sensing image and is still an image itself. The terrain parameters obtained in this way are more in line with the actual situation. Correspondingly, the input of the evaluation model also becomes an image format.

[0114] Figure 2 It is a block diagram of the composition structure of the anti-erosion and fertility improvement effect evaluation system for degraded black soil areas. In an embodiment of the present invention, an anti-erosion and fertility improvement effect evaluation system for degraded black soil areas, the system 10 includes:

[0115] An evaluation level determination module 11, configured to receive the improvement behavior input by the user and determine the evaluation level of the improvement behavior according to historical behaviors;

[0116] A soil data acquisition module 12, configured to select sampling points in the black soil improvement area and acquire soil data with time tags at the sampling points;

[0117] A data application module 13, configured to identify the soil data with time tags to generate an improvement effect and synchronously adjust the data acquisition frequency of the sampling points; the data acquisition frequency is proportional to the data weight at the sampling points, and the data weight is used to characterize the degree of difference between the data and other data; the minimum value of the data acquisition frequency is determined by the evaluation level;

[0118] A model training module 14, configured to obtain the terrain parameters at each sampling point, construct a sample set of terrain parameters and improvement effects under the improvement behavior, train the neural network model, and when the error rate of the neural network model is less than a preset error rate threshold, output the trained neural network model as an evaluation model.

[0119] Further, the evaluation level determination module 11 includes:

[0120] The behavior recording unit is configured to receive the improved behaviors input by the user; the improved behaviors are in the form of an array, and the serial number of each element in the array represents the operation item to be executed, and the value of the element is a number; the number contains sub-elements representing the operation type and sub-elements representing the operation amount;

[0121] The behavior statistics unit is configured to count each improvement behavior of each black soil improvement area to obtain historical behaviors; among them, the improvement behaviors in the historical behaviors contain time tags;

[0122] The comparison and application unit is configured to compare the improved behaviors input by the user with the historical behaviors to determine the evaluation level of the improved behaviors;

[0123] The process of determining the evaluation level is as follows:

[0124] ; where, is the evaluation level, is a preset gradient function, represents the independent variable, and the threshold value in the gradient function is a preset value, represents the number of behaviors in the historical behaviors whose similarity to the improved behaviors reaches the preset similarity threshold.

[0125] Specifically, the soil data acquisition module 12 includes:

[0126] The sampling area determination unit is configured to query the crop planting positions in the black soil improvement area and determine the sampling area based on the crop planting positions;

[0127] The sampling point selection unit is configured to randomly select a preset number of points in the sampling area as sampling points;

[0128] The acquisition unit is configured to install a data collector at the sampling points to acquire soil data; the soil data includes at least the visual information of the crops and the physical and chemical data of the soil.

[0129] Furthermore, the data application module 13 includes:

[0130] The data sorting unit is configured to sort any type of soil data according to the time tag to obtain a data sequence;

[0131] The data regularization unit is configured to regularize the data sequence based on a preset standard data to obtain a regularized sequence;

[0132] The difference calculation unit is configured to obtain the difference sequence of the regularized sequence and generate an improvement effect; the improvement effect is a parameter jointly determined by the regularized sequence and the difference sequence;

[0133] A frequency adjustment unit is used to compare all regular sequences at different sampling points and synchronously adjust the data acquisition frequency of the sampling points according to the comparison results; wherein, the comparison process of the regular sequences occurs between soil data of the same type.

[0134] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An evaluation method for the improvement effect of erosion resistance and soil fertility improvement in degenerated black soil areas, characterized in that, The method includes: Receiving the improved behavior input by the user, and determining the evaluation level of the improved behavior according to the historical behavior; Selecting sampling points in the black soil improvement area, and obtaining soil data with time tags at the sampling points; Identifying the soil data with time tags to generate an improvement effect, and synchronously adjusting the data acquisition frequency of the sampling points; the data acquisition frequency is proportional to the data weight at the sampling point, and the data weight is used to characterize the difference degree between the data and other data; the minimum value of the data acquisition frequency is determined by the evaluation level; Obtaining the terrain parameters at each sampling point, constructing a sample set of the terrain parameters and the improvement effect under the improved behavior, training a neural network model, and when the error rate of the neural network model is less than the preset error rate threshold, outputting the trained neural network model as an evaluation model.

2. The evaluation method for the anti-erosion and fertility improvement effect in degraded black soil areas according to claim 1, wherein, The step of receiving the improved behavior input by the user and determining the evaluation level of the improved behavior according to the historical behavior includes: Receiving the improved behavior input by the user; the improved behavior is in the form of an array, and the serial number of each element in the array represents the nth operation executed, and the value of the element uses a number; the number contains a sub-element representing the operation type and a sub-element representing the operation amount; Counting each improved behavior in each black soil improvement area to obtain the historical behavior; among them, the improved behavior in the historical behavior contains time tags; Comparing the improved behavior input by the user with the historical behavior to determine the evaluation level of the improved behavior; The process of determining the evaluation level is: ; where, is the evaluation level, is the preset gradient function, represents the independent variable, and the threshold value in the gradient function is a preset value, represents the number of behaviors in the historical behaviors whose similarity to the improvement behavior reaches the preset similarity threshold.

3. The evaluation method for the anti-erosion and fertility improvement effect in degenerated black soil areas according to claim 1, wherein The step of selecting sampling points in the black soil improvement area and obtaining soil data with time tags at the sampling points includes: Querying the crop planting positions in the black soil improvement area, and determining the sampling area based on the crop planting positions; Randomly selecting a preset number of points in the sampling area as sampling points; Installing a data collector at the sampling point to obtain soil data; the soil data at least includes the visual information of the crop and the physical and chemical data of the soil.

4. The evaluation method for the anti-erosion and fertility improvement effect in degenerated black soil areas according to claim 1, characterized in that The step of identifying the soil data with time tags to generate an improvement effect and synchronously adjusting the data acquisition frequency of the sampling points includes: Sorting any type of soil data according to the time tag to obtain a data sequence; Regularizing the data sequence based on the preset standard data to obtain a regularized sequence; Obtaining the difference sequence of the regularized sequence to generate an improvement effect; the improvement effect is a parameter jointly determined by the regularized sequence and the difference sequence; Comparing all the regularized sequences of different sampling points, and synchronously adjusting the data acquisition frequency of the sampling points according to the comparison result; among them, the comparison process of the regularized sequences occurs between the same type of soil data.

5. The evaluation method for the anti-erosion and fertility improvement effect in degenerated black soil areas according to claim 4, characterized in that, The step of comparing all the regularized sequences of different sampling points and synchronously adjusting the data acquisition frequency of the sampling points according to the comparison result includes: For any sampling point, sequentially determining the type of soil data; Reading the regularized sequence of this sampling point and the regularized sequences of other sampling points under this type, comparing the regularized sequences, and calculating the similarity and its similarity mean value; Counting the similarity mean values corresponding to each type to determine the data acquisition frequency of the sampling points; The process of determining the data acquisition frequency is: ; wherein, represents the data acquisition frequency, is a preset constant, is the total number of types, is the preset correction coefficient of the th type, is the average similarity of the 6. The evaluation method for the anti-erosion and fertility improvement effect in the degraded black soil area according to claim 3, wherein The steps of obtaining the terrain parameters at each sampling point, constructing a sample set of terrain parameters and improvement effects under the improvement behavior, training a neural network model, and outputting the trained neural network model as an evaluation model when the error rate of the neural network model is less than a preset error rate threshold include: Obtaining the remote sensing data of the sampling area corresponding to each sampling point based on the remote sensing data; Identifying the remote sensing data and extracting the terrain contour map as the terrain parameter; Reading the improvement effect of the sampling point and establishing a sample set from the terrain parameter to the improvement effect; Splitting the sample set into a training set, a test set and a validation set, and training the neural network model; When the error rate of the neural network model is less than a preset error rate threshold, outputting the trained neural network model as an evaluation model.

7. An evaluation system for the improvement effect of erosion resistance and soil fertility enhancement in degenerated black soil areas, characterized in that, The system includes: An evaluation level determination module for receiving the improvement behavior input by the user and determining the evaluation level of the improvement behavior according to the historical behavior; A soil data acquisition module for selecting sampling points in the black soil improvement area and acquiring soil data with time tags at the sampling points; A data application module for identifying the soil data with time tags to generate an improvement effect and synchronously adjusting the data acquisition frequency of the sampling points; the data acquisition frequency is proportional to the data weight at the sampling point, and the data weight is used to characterize the difference degree between the data and other data; the minimum value of the data acquisition frequency is determined by the evaluation level; A model training module for obtaining the terrain parameters at each sampling point, constructing a sample set of terrain parameters and improvement effects under the improvement behavior, training a neural network model, and outputting the trained neural network model as an evaluation model when the error rate of the neural network model is less than a preset error rate threshold.

8. The anti-erosion and fertility improvement effect evaluation system for degraded black soil areas according to claim 7, characterized in that The evaluation level determination module includes: A behavior recording unit for receiving the improvement behavior input by the user; the improvement behavior is in the form of an array, and the serial number of each element in the array represents the operation item executed, and the value of the element is a number; the number contains a sub-element representing the operation type and a sub-element representing the operation amount; A behavior statistics unit for statistically analyzing each improvement behavior of each black soil improvement area to obtain the historical behavior; among them, the improvement behavior in the historical behavior contains time tags; A comparison and application unit for comparing the improvement behavior input by the user with the historical behavior to determine the evaluation level of the improvement behavior; The process of determining the evaluation level is: ; wherein, is the evaluation level, is the preset gradient function, represents the independent variable, and the threshold value in the gradient function is a preset value, represents the number of behaviors in the historical behaviors whose similarity to the improved behavior reaches the preset similarity threshold.

9. The evaluation system for the anti-erosion and fertility improvement effect in the degraded black soil area according to claim 7, characterized in that, The soil data acquisition module includes: A sampling area determination unit for querying the crop planting positions in the black soil improvement area and determining the sampling area based on the crop planting positions; A sampling point selection unit for randomly selecting a preset number of points in the sampling area as sampling points; An acquisition unit for installing a data collector at the sampling point to acquire soil data; the soil data at least includes the visual information of the crop and the physical and chemical data of the soil.

10. The evaluation system for anti-erosion and fertility improvement effects in degraded black soil areas according to claim 7, characterized in that, The data application module includes: A data sorting unit for sorting any type of soil data according to the time tag to obtain a data sequence; A data regularization unit for regularizing the data sequence based on a preset standard data to obtain a regularized sequence; Differential calculation unit, configured to obtain a differential sequence of a regular sequence and generate an improvement effect; the improvement effect is a parameter jointly determined by the regular sequence and the differential sequence; Frequency adjustment unit, configured to compare all regular sequences at different sampling points and synchronously adjust the data acquisition frequency of the sampling points according to the comparison result; wherein, the comparison process of the regular sequences occurs between soil data of the same type.

Images