An improved method and improved system based on salinity analysis of saline-alkali land

Abstract

The present invention belongs to the technical field of saline-alkali land improvement, and discloses an improvement method and improvement system based on saline-alkali land salinity analysis. The method includes: collecting environmental data of the target farming area and soil moisture data at the sampling point in real time, and summarizing into a data set in combination with the sampling point location information; constructing an online real-time monitoring model, with environmental data, soil temperature / humidity and groundwater level as input, and soil conductivity EC and soil pH as output; filling the soil temperature / humidity and groundwater level at non-sampling points with data; using the online real-time monitoring model, with the environmental data collected in real time and the filling data as input, predicting the soil conductivity EC and soil pH at non-sampling points; integrating the sampling point and non-sampling point data to obtain an online monitoring map of soil salinity status, and realizing real-time online monitoring and display of soil moisture data and environmental data at any point in the target farming area; locating the area to be improved according to the online monitoring map of soil salinity status, matching the improvement strategy, and ensuring accurate and efficient improvement.

Description

Technical Field

[0001] The present invention belongs to the technical field of saline-alkali land improvement, and in particular relates to an improvement method and an improvement system based on saline-alkali land salinity analysis. Background Art

[0002] Saline-alkali land is a general term for saline soil, alkaline soil and other soils with different degrees of salinization and alkalization. It has high salt content or alkalinity, poor soil structure, easy compaction, low organic matter content, poor nutrients, and weak soil fertilizer retention capacity, which affects crop growth. It is a kind of soil with poor properties and low fertility. Saline-alkali land is usually formed in areas with arid climate and water evaporation rate greater than water replenishment rate, such as coastal areas, inland basins and desert areas. There are two main reasons for the formation of saline-alkali land: salt accumulation and alkaline substance accumulation. In arid areas, the water evaporation rate is greater than the water replenishment rate, resulting in the inability to fully wash the salt in the soil, thus accumulating in the soil. At the same time, alkaline substances in groundwater may also enter the soil through leakage and soaking, resulting in soil alkalinization and increased pH value.

[0003] At present, saline-alkali land is widely distributed in my country and is an important reserve resource for arable land. The development, management and improvement of saline-alkali land to form arable land is of great significance for increasing grain production, improving land resource utilization and promoting agricultural development.

[0004] The existing patent CN116703181A discloses a method for improving saline-alkali land, which comprehensively generates the saline-alkali condition coefficient of the saline-alkali land area to be treated, and sends a request message to the outside when it is greater than the preset condition threshold; generates the saline-alkali state coefficient from the acquired saline-alkali state data set, determines the alkalization area and the non-alkalization area; corrects the saline-alkali state coefficient and divides the detection area, and selects the corresponding treatment plan according to the division result; establishes a saline-alkali feature library and a saline-alkali land treatment plan library, and matches the corresponding saline-alkali land treatment plan according to the correspondence between the saline-alkali characteristics of the saline-alkali area and the saline-alkali land treatment plan; uses the saline-alkali land treatment prediction model to predict and analyze the saline-alkali land treatment plan output by the matching output, and selects the saline-alkali land treatment plan according to the prediction result. Matching the corresponding treatment plan avoids the situation where the selected treatment plan does not correspond to the actual situation, and maintains the effectiveness of saline-alkali land treatment. However, the technical solution of this patent only manages saline-alkali land in different regions according to the salinity condition coefficient determined by the groundwater level and the average annual precipitation, ignoring the detection and management of non-alkalinized areas; and the electrical conductivity EC, salinity stress index Esp and sodium adsorption ratio Sar in the salinity area are collected at the detection point, and a salinity status data set is established after aggregation, without considering the influence of external environmental factors such as light, temperature and humidity, and crop growth time nodes on the salinity status of the soil; at the same time, drone inspections and low-altitude imaging are required during the management process to obtain image information of the alkalinized area, and real-time monitoring and prediction of the salinity status of the land cannot be achieved. Summary of the invention

[0005] The present invention aims to address the technical problems existing in the background technology and provide an improvement method and an improvement system based on salinity analysis of saline-alkali land, which can combine external environmental data and soil moisture data to monitor and predict the soil salinity status of the entire target farming area and the soil salinity status of the entire growth cycle of crops online in real time, accurately grasp the changes in indicators at each position in the target farming area, so as to timely and scientifically match soil improvement strategies, improve saline-alkali land soil more accurately, efficiently and with high quality, and provide customized solutions for the comprehensive management of saline-alkali land.

[0006] In order to achieve the above technical objectives, the present invention adopts the following technical solutions:

[0007] An improved method for analyzing salinity in saline-alkali land comprises the following steps:

[0008] Step S1: Data collection, including:

[0009] Obtain an electronic map of the target farming area; collect environmental data of the target farming area and soil moisture data at the sampling points in real time; and aggregate the collected data and the sampling point location information into a cloud platform database to form a data set;

[0010] Step S2: Building an online real-time monitoring model, including:

[0011] Using a neural network model, the data set formed by sampling at the sampling points is divided into a training set and a test set, and the neural network model is trained and optimized. The optimized model is the online real-time monitoring model;

[0012] The input of the neural network model is: atmospheric temperature, atmospheric humidity, light intensity, rainfall, soil temperature, soil moisture and groundwater level; the output is: soil electrical conductivity EC and soil pH value;

[0013] Step S3: data filling of soil temperature, soil moisture and groundwater level data at non-sampling points;

[0014] Step S4: using the online real-time monitoring model, taking the real-time collected atmospheric temperature, atmospheric humidity, light intensity and rainfall, and the soil temperature, soil moisture and groundwater level data at the non-sampling points filled in step S3 as input, predicting the soil electrical conductivity EC and soil pH value at the non-sampling points;

[0015] Step S5: Integrate the soil moisture data collected in real time at the sampling point with the soil moisture data at the non-sampling point obtained in steps S3-S4, and combine them with the environmental data of the target farming area collected in real time, so as to obtain an online monitoring map of the soil salinity state of the target farming area, and realize real-time online monitoring and display of the soil moisture data and environmental data at each position of the target farming area;

[0016] Step S6: Based on the real-time monitoring results of the online monitoring map of the soil salinity and alkali status of the target farming area, locate the area to be improved, and match the improvement strategy in combination with the environmental data and soil moisture data.

[0017] Further, in step S1, the environmental data includes: atmospheric temperature, atmospheric humidity, light intensity and rainfall;

[0018] The soil moisture data include: soil electrical conductivity EC, soil pH value, soil temperature, soil moisture and groundwater level.

[0019] Furthermore, in step S1, the target farming area is divided into grids, and soil moisture sampling points are arranged in the center of the grids. At the sampling points, sensors monitor and collect soil moisture data in real time and upload the data every 5-10 minutes.

[0020] Furthermore, in step S3, the Kriging interpolation method is used to fill in the soil temperature, soil moisture and groundwater level data at non-sampling points.

[0021] Furthermore, the step S6 specifically includes:

[0022] Step S61: determining whether the soil electrical conductivity EC and soil pH value at a certain point are higher than a preset condition threshold;

[0023] Step S62: when any one of the soil electrical conductivity EC and soil pH value is higher than a preset condition threshold, the point is marked as a data abnormal point;

[0024] Step S63: defining a circular area with the location of the data abnormal point as the center and the set distance as the radius, and determining the number of data abnormal points in the circular area;

[0025] Step S64: When the number of abnormal data points is greater than a preset number, an early warning is initiated, and the circular area is used as an area to be improved for early warning prompts;

[0026] Step S65: For the area to be improved, the corresponding environmental data and soil moisture data are combined to match and output the corresponding improvement strategy.

[0027] Furthermore, the method further comprises the following steps:

[0028] Step S7: Regularly calibrate and update the online real-time monitoring model in step S2 based on the data collected at the sampling points.

[0029] Furthermore, the method further comprises the following steps:

[0030] Step S8: construct a soil salinity prediction model to predict the soil salinity of the entire target farming area for the next n days, including:

[0031] Step S81: Based on the historical collected data consisting of the environmental data and soil moisture data collected at the sampling point, a neural network model is used, with the environmental data and soil moisture data of the sampling point N days before and the environmental data of the N+1 day as input, and the soil electrical conductivity EC and soil pH value of the N+1 day as output, to predict the soil salinity of the sampling point on the N+1 day; N=5-7;

[0032] The historical data collected at the sampling points are processed, and the historical data are divided into several sample sets in units of N+1 days. The model is trained, tested and optimized to obtain a soil salinity prediction model.

[0033] Step S82: According to the soil salinity prediction model, using the data collected at the sampling points of the current N days, predict the soil electrical conductivity EC and soil pH value at the sampling points of the next n days in sequence, where n=1-3;

[0034] Step S83: According to the soil electrical conductivity EC and soil pH value at the sampling points for the next n days predicted in step S82, as well as the currently collected soil temperature, soil moisture and groundwater level as the soil temperature, soil moisture and groundwater level for the next n days, and the environmental data obtained by the weather forecast for the next n days, the method of steps S3-S5 is repeated to predict the soil salinity and alkali status of the entire target farming area for the next n days.

[0035] Furthermore, in step S82:

[0036] When n=1, the model predicts the soil electrical conductivity EC and soil pH value at the sampling point in the next day by using the data collected at the sampling point in the current N days and combining it with the environmental data obtained from the weather forecast in the next day.

[0037] When n=2, the data collected at the sampling point on the current N-1 day are used, combined with the environmental data obtained from the weather forecast for the next two days, the soil conductivity EC and soil pH value at the sampling point on the next day predicted by the model, and the soil temperature, soil moisture and groundwater level collected at the current time are used as the soil temperature, soil moisture and groundwater level on the first day of the future. The model predicts the soil conductivity EC and soil pH value at the sampling point on the next two days.

[0038] When n=3, the collected data at the sampling point of the current N-2 days are used, combined with the environmental data obtained from the weather forecast for the next 3 days, the soil electrical conductivity EC and soil pH values ​​at the sampling points for the next 2 days predicted by the model, and the currently collected soil temperature, soil moisture and groundwater level are used as the soil temperature, soil moisture and groundwater level for the next 2 days. The model predicts the soil electrical conductivity EC and soil pH values ​​at the sampling points for the next 3 days.

[0039] At the same time, the present invention also provides an improved system based on salinity analysis of saline-alkali land, which is constructed using the improved method described in any one of the above items, comprising:

[0040] Data acquisition module: used to acquire electronic maps of target farming areas, land environment data, soil moisture data, and weather forecast data of target farming areas;

[0041] Online real-time monitoring module: It has a built-in online real-time monitoring model; the online real-time monitoring model uses the real-time collected environmental data, as well as the soil temperature, soil moisture and groundwater level data at non-sampling points as input, predicts the soil conductivity EC and soil pH at non-sampling points, and integrates them with the soil moisture data collected at the collection points to obtain an online monitoring map of the soil salinity and alkali status of the target farming area, and realizes real-time online monitoring and real-time display of soil moisture and environmental data at each location in the target farming area;

[0042] Strategy matching module: It is used to locate the area to be improved based on the results of the online monitoring map of soil salinity and alkali status, match the improvement strategy for the area to be improved, and output soil improvement suggestions.

[0043] Furthermore, it also includes:

[0044] Soil salinity prediction module: Based on the historical data collected at the sampling point, which consists of environmental data and soil moisture data, and using the data collected in the previous N days, predict the soil conductivity EC and soil pH value that characterize the soil salinity at the sampling point in the next N days. Combined with the currently collected soil temperature, soil moisture and groundwater level as the soil temperature, soil moisture and groundwater level in the next n days, as well as the environmental data obtained by the weather forecast in the next n days, the online real-time monitoring model constructed in the online real-time monitoring module is used to predict the soil salinity of the entire target farming area in the next n days. The forecast result map, so that operators can predict the trend of soil salinization and guide improvement strategies;

[0045] Update module: Based on the collected data at the sampling points, the online real-time monitoring model is updated regularly, and the soil salinity prediction model is continuously updated to improve the model prediction accuracy, so as to more accurately match the improvement strategy and ensure the improvement effect.

[0046] Compared with the prior art, the beneficial effects of the present invention are:

[0047] (1) The improved method and improved system based on salinity analysis of saline-alkali land provided by the present invention use the data set collected at the sampling point to train and construct an online real-time monitoring model, use the Kriging interpolation data to fill in the soil temperature, soil moisture and groundwater level data at the non-sampling points, and then use the online real-time monitoring model to use the real-time collected environmental data and the filled soil temperature, soil moisture and groundwater level data at the non-sampling points as input to predict the soil electrical conductivity EC and soil pH value at the non-sampling points, integrate and obtain the online monitoring map of the soil salinity status of the target farming area, and realize the soil electrical conductivity EC and soil pH value at each position in the target farming area. Real-time online monitoring of soil pH value, and the simultaneous display of atmospheric temperature, atmospheric humidity, light intensity, rainfall, soil temperature, soil moisture and groundwater level at any location in the target farming area, realizes online real-time monitoring and prediction of the soil salinity and alkali status of the entire target farming area and the soil salinity and alkali status of the entire crop growth cycle, accurately grasps the changes in indicators at each location in the target farming area, so as to timely and scientifically match soil improvement strategies, and improve saline-alkali land soil more accurately, efficiently and with high quality. Especially when conducting supervision of large-scale target farming areas, it can provide customized solutions for the comprehensive management of saline-alkali land, significantly reducing costs and increasing efficiency;

[0048] (2) According to the real-time monitoring results of the online monitoring map of the soil salinity status of the target farming area, when locating the position of the area to be improved, the data anomaly points are marked by the conditional threshold, and a circular area is delineated with the data anomaly point as the center and the set distance as the radius, and the number of data anomaly points in the circular area is determined; when the number of data anomaly points is greater than the preset number, an early warning is activated, and the circular area is highlighted as the area to be improved; for the area to be improved, the environmental data and soil moisture data here are combined to match and output the corresponding improvement strategy; this early warning method can reduce the occurrence of misjudgment caused by model prediction errors, accurately realize targeted soil improvement, and thus ensure the improvement effect;

[0049] (3) The embodiment of the present invention predicts the soil salinity of the next n days based on the sampling point data of the previous N days. This is a continuous prediction based on the consideration of the growth trend of crops, and the prediction result is relatively accurate. When the data for more than one day needs to be predicted, the currently collected data is used for filling. The calculation method is simple, and both the prediction accuracy and the calculation efficiency are taken into account. The future trend of soil salinization can be predicted, which is convenient for taking preventive and improvement measures in advance, reversing the trend of soil salinization and deterioration in advance, and ensuring the growth of crops. In addition, by comparing the collected data at the sampling points of the current n days with the prediction results of the next n days corresponding to the prediction model of the soil salinity, the model parameters of the soil salinity prediction model are continuously corrected and optimized, which can further improve the prediction results, more accurately realize the monitoring and management of the entire life cycle of crops, and achieve scientific improvement and efficient farmland cultivation. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 The overall flow chart of the improved method for analyzing salinity in saline-alkali land according to an embodiment of the present invention;

[0051] Figure 2 is a flow chart of step S6 of an embodiment of the present invention;

[0052] Figure 3 Flow chart of step S8 of the embodiment of the present invention. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] Example 1

[0055] Combination Figure 1As shown, an embodiment of the present invention provides an improved method for analyzing salinity in saline-alkali land, and the improved method comprises the following steps:

[0056] Step S1: Data collection, including:

[0057] Obtain an electronic map of the target farming area; collect environmental data of the target farming area and soil moisture data at the sampling points in real time; and aggregate the collected data together with the sampling point location information into a cloud platform database to form a data set.

[0058] In the above step S1, the environmental data include: atmospheric temperature, atmospheric humidity, light intensity and rainfall; since the environmental data of the same farming area are basically the same, only one set of environmental data collection device is required for the target farming area.

[0059] The soil moisture data include: soil electrical conductivity EC, soil pH value, soil temperature, soil moisture and groundwater level; specifically, the target farming area can be divided into grids, and soil moisture sampling points can be arranged in the center of the grid. At the sampling points, sensors monitor and collect soil moisture data in real time and upload it every 5-10 minutes.

[0060] Step S2: Building an online real-time monitoring model, including:

[0061] Using a neural network model, the data set formed by sampling at the sampling points is divided into a training set and a test set, and the neural network model is trained and optimized. The optimized model is the online real-time monitoring model;

[0062] The input of the neural network model is: atmospheric temperature, atmospheric humidity, light intensity, rainfall, soil temperature, soil moisture and groundwater level; the output is: soil electrical conductivity EC and soil pH value;

[0063] In step S2, the neural network model may specifically be a BP neural network model.

[0064] Step S3: Using Kriging interpolation method, the data are filled with soil temperature, soil moisture and groundwater level data at non-sampling points;

[0065] Since soil temperature, soil moisture and groundwater level have a certain correlation in space, the Kriging interpolation method is used here to fill in the data. Filling in the relevant data based on spatial autocorrelation can be closer to the true value, which is conducive to improving the prediction accuracy of the subsequent model.

[0066] Step S4: Using the online real-time monitoring model, the real-time collected atmospheric temperature, atmospheric humidity, light intensity and rainfall, as well as the soil temperature, soil moisture and groundwater level data at the non-sampling points filled in step S3 are used as input to predict the soil electrical conductivity EC and soil pH at the non-sampling points.

[0067] Step S5: Integrate the soil moisture data collected in real time at the sampling point with the soil moisture data at the non-sampling point obtained in steps S3-S4, and combine them with the real-time collected environmental data of the target farming area, so as to obtain an online monitoring map of the soil salinity status of the target farming area, and realize real-time online monitoring and display of soil moisture data and environmental data at each location in the target farming area.

[0068] Specifically, the soil conductivity EC and soil pH value collected in real time at the sampling point are integrated with the soil conductivity EC and soil pH value at the non-sampling point predicted and obtained in step S4 to obtain an online monitoring diagram of the soil salinity state of the target farming area, so as to realize the real-time online monitoring of the soil conductivity EC and soil pH value at each position of the target farming area, and simultaneously display the atmospheric temperature, atmospheric humidity, light intensity, rainfall, soil temperature, soil humidity and groundwater level at any position of the target farming area. For non-sampling points, the displayed soil moisture data is the data obtained by filling, and for sampling points, the displayed soil moisture data is the measured data; for a single target farming area, the same environmental data collected in real time can be shared.

[0069] Step S6: According to the real-time monitoring results of the online monitoring map of the soil salinity and alkali status of the target farming area, the location of the area to be improved is located, and the improvement strategy is matched in combination with the environmental data and the soil moisture data; specifically, the steps include:

[0070] Step S61: determining whether the soil electrical conductivity EC and soil pH value at a certain point are higher than a preset condition threshold;

[0071] Step S62: When any one of the soil electrical conductivity EC and the soil pH value is higher than a preset condition threshold (e.g., soil electrical conductivity EC>1.5mS / cm or pH>7.5), the point is marked as a data abnormal point;

[0072] Step S63: taking the location of the data abnormal point as the center and defining a circular area with a set distance (such as 5 meters) as the radius, and determining the number of data abnormal points in the circular area;

[0073] Step S64: When the number of abnormal data points is greater than a preset number (such as 100), an early warning is initiated, and the circular area is used as an area to be improved for early warning (such as color differentiation display or highlight flashing or prompting early warning information, etc.);

[0074] Step S65: For the area to be improved, the environmental data and soil moisture data there are combined to match and output the corresponding improvement strategy.

[0075] For example: if the soil electrical conductivity EC of the area to be improved is high, it means that the soil salinity is high, and irrigation can be used to wash away the salt; if the soil pH value of the area to be improved is high, it means that the soil alkalinity is high, which can be improved by increasing the application of organic fertilizers or soil conditioners; if both values ​​are relatively high, comprehensive management measures should be matched, such as covering with grass, spreading sand, irrigation, fertilization and other measures should be carried out simultaneously.

[0076] The above step S6 can reduce the occurrence of misjudgments caused by model prediction errors, accurately achieve targeted soil improvement, and thus ensure the improvement effect.

[0077] In addition, the following steps are included:

[0078] Step S7: Regularly calibrate and update the online real-time monitoring model in step S2 based on the data collected at the sampling points to continuously improve the prediction accuracy of the entire method.

[0079] The improved method based on salinity analysis of saline-alkali land provided by the embodiment of the present invention uses the data set collected at the sampling point to train and construct an online real-time monitoring model, uses the Kriging interpolation data to fill in the soil temperature, soil moisture and groundwater level data at the non-sampling points, and then according to the online real-time monitoring model, uses the real-time collected environmental data and the filled soil temperature, soil moisture and groundwater level data at the non-sampling points as input to predict the soil electrical conductivity EC and soil pH value at the non-sampling points, integrates and obtains the online monitoring map of the soil salinity state of the target farming area, and realizes the soil electrical conductivity EC and soil pH value at each position in the target farming area. Real-time online monitoring of the H value can simultaneously display the atmospheric temperature, atmospheric humidity, light intensity, rainfall, soil temperature, soil moisture and groundwater level at any location in the target farming area, realizing online real-time monitoring and prediction of the soil salinity of the entire target farming area and the soil salinity of the entire growth cycle of crops, and accurately grasping the changes in indicators at each location in the target farming area, so as to timely and scientifically match the soil improvement strategy, and improve the saline-alkali land soil more accurately, efficiently and with high quality. Especially when supervising large-scale target farming areas, it can provide customized solutions for the comprehensive management of saline-alkali land, significantly reducing costs and increasing efficiency.

[0080] Example 2

[0081] The improved method of the embodiment of the present invention is further improved on the basis of embodiment 1.

[0082] The improved method of the embodiment of the present invention further comprises the following steps after steps S1 to S7 of embodiment 1:

[0083] Step S8: constructing a soil salinity prediction model to predict the soil salinity of the entire target farming area for the next n days; specifically including:

[0084] Step S81: Based on the historical collected data consisting of the environmental data and soil moisture data collected at the sampling point, a neural network model is used, with the environmental data and soil moisture data of the sampling point N days before and the environmental data of the N+1 day as input, and the soil electrical conductivity EC and soil pH value of the N+1 day as output, to predict the soil salinity of the sampling point on the N+1 day; N=5-7;

[0085] The historical data collected at the sampling points are processed, and the historical data are divided into several sample sets in units of N+1 days. The model is trained, tested and optimized to obtain a soil salinity prediction model.

[0086] As one example, a neural network model is used to predict the soil electrical conductivity EC and soil pH value characterizing the soil salinity state in the next day (i.e., the 8th day) by using the collected data of the sampling points in the previous 7 days. The historical collected data of the sampling points can be processed, and the historical collected data can be divided into several sample sets in units of 8 days. The model is trained, tested and optimized to obtain a soil salinity state prediction model; and, in the same way, the days are postponed one day in sequence, and the soil salinity state in the next 2 days (i.e., the 9th day), the next 3 days (i.e., the 10th day), and so on can be predicted; because the more the prediction data used in the input variables are postponed, the worse the prediction accuracy is, so n=1-3 days is preferred here.

[0087] Step S82: According to the soil salinity prediction model, using the data collected at the sampling points of the current N days (including the current day), predict the soil electrical conductivity EC and soil pH value at the sampling points of the next n days in sequence; n = 1-3;

[0088] When n=1, the model predicts the soil electrical conductivity EC and soil pH value at the sampling point in the next day by using the data collected at the sampling point in the current N days (including the current day) (including environmental data and soil moisture data) and combining it with the environmental data obtained from the weather forecast for the next day;

[0089] When n=2, the collected data (including environmental data and soil moisture data) at the sampling point of the current N-1 days (including the current day) are used, combined with the environmental data obtained by the weather forecast for the next two days, the soil conductivity EC and soil pH value at the sampling point of the next day predicted by the model, and the soil temperature, soil moisture and groundwater level currently collected are used as the soil temperature, soil moisture and groundwater level of the first day in the future, and the model predicts the soil conductivity EC and soil pH value at the sampling point of the next two days;

[0090] When n=3, the collected data (including environmental data and soil moisture data) at the sampling points of the current N-2 days (including the current day) are used, combined with the environmental data obtained by the weather forecast for the next 3 days, the soil conductivity EC and soil pH values ​​at the sampling points for the next 2 days obtained by the model prediction, and the soil temperature, soil moisture and groundwater level currently collected are used as the soil temperature, soil moisture and groundwater level for the next 2 days, and the model predicts the soil conductivity EC and soil pH values ​​at the sampling points for the next 3 days;

[0091] Step S83: According to the soil electrical conductivity EC and soil pH value at the sampling points for the next n days predicted in step S82, as well as the currently collected soil temperature, soil moisture and groundwater level as the soil temperature, soil moisture and groundwater level for the next n days, and the environmental data obtained by the weather forecast for the next n days, the method of steps S3-S5 is repeated to predict the soil salinity and alkali status of the entire target farming area for the next n days.

[0092] As a further improvement, the step S8 further includes the following steps:

[0093] Step S84: Compare the collected data at the sampling points of the current n days (including the current day) with the prediction results of the soil salinity prediction model for the next n days, and continuously correct and optimize the model parameters of the soil salinity prediction model to improve the prediction results.

[0094] The embodiment of the present invention predicts the soil salinity status of the next n days based on the sampling point data of the current N days. It is a continuous prediction based on the consideration of the growth trend of crops, and the prediction result is relatively accurate. When the data for more than 1 day needs to be predicted, the currently collected data is used for filling. The calculation method is simple, and the prediction accuracy and calculation efficiency are taken into account. It can predict the future trend of soil salinization, facilitate the adoption of preventive and improvement measures in advance, reverse the trend of soil salinization and deterioration in advance, and ensure the growth of crops.

[0095] In addition, by comparing the collected data at the sampling points on the current n days with the prediction results of the soil salinity status prediction model for the next n days, and continuously correcting and optimizing the model parameters of the soil salinity status prediction model, the prediction results can be further improved, and more accurate monitoring and management of the entire life cycle of crops can be achieved, thereby realizing scientific improvement and efficient field cultivation.

[0096] Example 3

[0097] The embodiment of the present invention provides an improved system based on salinity analysis of saline-alkali land, which is constructed using the improved method described in Examples 1-2, and includes:

[0098] Data acquisition module: used to acquire electronic maps of target farming areas, land environment data, soil moisture data, and weather forecast data of target farming areas;

[0099] Online real-time monitoring module: It has a built-in online real-time monitoring model; the online real-time monitoring model uses the real-time collected environmental data, as well as the soil temperature, soil moisture and groundwater level data at non-sampling points as input, predicts the soil conductivity EC and soil pH at non-sampling points, and integrates them with the soil moisture data collected at the collection points to obtain an online monitoring map of the soil salinity and alkali status of the target farming area, and realizes real-time online monitoring and real-time display of soil moisture and environmental data at each location in the target farming area;

[0100] Strategy matching module: used to locate the area to be improved based on the results of the online monitoring map of soil salinity and alkali status, match the improvement strategy for the area to be improved, and output soil improvement suggestions;

[0101] Soil salinity prediction module: Based on the historical collected data consisting of environmental data and soil moisture data collected at the sampling points, the collected data of the previous N days are used to predict the soil electrical conductivity EC and soil pH value that characterize the soil salinity at the sampling points in the next n days. The currently collected soil temperature, soil moisture and groundwater level are used as the soil temperature, soil moisture and groundwater level for the next n days, as well as the environmental data obtained by the weather forecast for the next n days. The online real-time monitoring model constructed in the online real-time monitoring module is used to predict the soil salinity of the entire target farming area in the next n days. The prediction result map, so that the operating personnel can predict the trend of soil salinization and guide the improvement strategy.

[0102] In addition, in some embodiments, the improved system further comprises:

[0103] Update module: Based on the collected data at the sampling points, the online real-time monitoring model is updated regularly, and the soil salinity prediction model is continuously updated to improve the model prediction accuracy, so as to more accurately match the improvement strategy and ensure the improvement effect.

[0104] The above description is only an embodiment of the present invention and is not intended to limit the present invention. Any modification, equivalent replacement and improvement made within the application scope of the present invention shall be included in the protection scope of the present invention.

Claims

1. An improved method for analyzing salinity in saline-alkali land, characterized in that: The improved method comprises the following steps: Step S1: Data collection, including: Obtain an electronic map of the target farming area; collect environmental data of the target farming area and soil moisture data at the sampling point in real time; aggregate the collected data and the sampling point location information into a cloud platform database to form a data set; wherein the soil moisture data includes: soil electrical conductivity EC, soil pH value, soil temperature, soil moisture and groundwater level; Step S2: Building an online real-time monitoring model, including: Using a neural network model, the data set formed by sampling at the sampling points is divided into a training set and a test set, and the neural network model is trained and optimized. The optimized model is the online real-time monitoring model; The input of the neural network model is: atmospheric temperature, atmospheric humidity, light intensity, rainfall, soil temperature, soil moisture and groundwater level; the output is: soil electrical conductivity EC and soil pH value; Step S3: data filling of soil temperature, soil moisture and groundwater level data at non-sampling points; Step S4: using the online real-time monitoring model, taking the real-time collected atmospheric temperature, atmospheric humidity, light intensity and rainfall, and the soil temperature, soil moisture and groundwater level data at the non-sampling points filled in step S3 as input, predicting the soil electrical conductivity EC and soil pH value at the non-sampling points; Step S5: Integrate the soil moisture data collected in real time at the sampling point with the soil moisture data at the non-sampling point obtained in steps S3-S4, and combine them with the environmental data of the target farming area collected in real time, so as to obtain an online monitoring map of the soil salinity state of the target farming area, and realize real-time online monitoring and display of the soil moisture data and environmental data at each position of the target farming area; Step S6: Based on the real-time monitoring results of the online monitoring map of the soil salinity and alkali status of the target farming area, locate the area to be improved, and match the improvement strategy in combination with the environmental data and soil moisture data.

2. The improved method for analyzing salinity in saline-alkali land according to claim 1, characterized in that: In step S1, the environmental data includes: atmospheric temperature, atmospheric humidity, light intensity and rainfall.

3. The improved method based on salinity analysis of saline-alkali land according to claim 2, characterized in that: In step S1, the target farming area is divided into grids, soil moisture sampling points are arranged at the center of the grids, and sensors are used to monitor and collect soil moisture data in real time at the sampling points, which are uploaded every 5-10 minutes.

4. The improved method for analyzing salinity in saline-alkali land according to claim 1, characterized in that: In step S3, the Kriging interpolation method is used to fill in the soil temperature, soil moisture and groundwater level data at non-sampling points.

5. The improved method for analyzing salinity in saline-alkali land according to claim 1, characterized in that: The step S6 specifically includes: Step S61: determining whether the soil electrical conductivity EC and soil pH value at a certain point are higher than a preset condition threshold; Step S62: when any one of the soil electrical conductivity EC and soil pH value is higher than a preset condition threshold, the point is marked as a data abnormal point; Step S63: defining a circular area with the location of the data abnormal point as the center and the set distance as the radius, and determining the number of data abnormal points in the circular area; Step S64: When the number of abnormal data points is greater than a preset number, an early warning is initiated, and the circular area is used as an area to be improved for early warning prompts; Step S65: For the area to be improved, the corresponding environmental data and soil moisture data are combined to match and output the corresponding improvement strategy.

6. The improved method for analyzing salinity in saline-alkali land according to claim 1, characterized in that: The following steps are also included: Step S7: Regularly calibrate and update the online real-time monitoring model in step S2 based on the data collected at the sampling points.

7. The improved method for analyzing salinity in saline-alkali land according to claim 1, characterized in that: The following steps are also included: Step S8: construct a soil salinity prediction model to predict the soil salinity of the entire target farming area for the next n days, including: Step S81: Based on the historical data consisting of environmental data and soil moisture data collected at the sampling point, a neural network model is used, with the environmental data and soil moisture data of the sampling point N days before and the environmental data of the N+1 day as input, and the soil electrical conductivity EC and soil pH value of the N+1 day as output, to predict the soil salinity of the sampling point on the N+1 day; N=5-7; The historical data collected at the sampling points are processed, and the historical data are divided into several sample sets in units of N+1 days. The model is trained, tested and optimized to obtain a soil salinity prediction model. Step S82: According to the soil salinity prediction model, using the data collected at the sampling points of the current N days, predict the soil electrical conductivity EC and soil pH value at the sampling points of the next n days in sequence, where n=1-3; Step S83: According to the soil electrical conductivity EC and soil pH value at the sampling points for the next n days predicted in step S82, as well as the currently collected soil temperature, soil moisture and groundwater level as the soil temperature, soil moisture and groundwater level for the next n days, and the environmental data obtained by the weather forecast for the next n days, the method of steps S3-S5 is repeated to predict the soil salinity and alkali status of the entire target farming area for the next n days.

8. The improved method for analyzing salinity in saline-alkali land according to claim 7, characterized in that: In step S82: When n=1, the model predicts the soil electrical conductivity EC and soil pH value at the sampling point in the next day by using the data collected at the sampling point in the current N days and combining it with the environmental data obtained from the weather forecast in the next day. When n=2, the data collected at the sampling point on the current N-1 day are used, combined with the environmental data obtained from the weather forecast for the next two days, the soil conductivity EC and soil pH values ​​at the sampling point on the next day predicted by the model, and the soil temperature, soil moisture and groundwater level collected at the current time are used as the soil temperature, soil moisture and groundwater level on the first day of the future. The model predicts the soil conductivity EC and soil pH values ​​at the sampling point on the next two days. When n=3, the collected data at the sampling point of the current N-2 days are used, combined with the environmental data obtained from the weather forecast for the next 3 days, the soil electrical conductivity EC and soil pH values ​​at the sampling points for the next 2 days predicted by the model, and the currently collected soil temperature, soil moisture and groundwater level are used as the soil temperature, soil moisture and groundwater level for the next 2 days. The model predicts the soil electrical conductivity EC and soil pH values ​​at the sampling points for the next 3 days.

9. An improvement system based on salinity analysis of saline-alkali land, constructed by the improvement method according to any one of claims 1 to 8, characterized in that: include: Data acquisition module: used to acquire electronic maps of target farming areas, land environment data, soil moisture data, and weather forecast data of target farming areas; Online real-time monitoring module: It has a built-in online real-time monitoring model; the online real-time monitoring model uses the real-time collected environmental data, as well as the soil temperature, soil moisture and groundwater level data at non-sampling points as input, predicts the soil conductivity EC and soil pH at non-sampling points, and integrates them with the soil moisture data collected at the collection points to obtain an online monitoring map of the soil salinity and alkali status of the target farming area, and realizes real-time online monitoring and real-time display of soil moisture and environmental data at each location in the target farming area; Strategy matching module: It is used to locate the area to be improved based on the results of the online monitoring map of soil salinity and alkali status, match the improvement strategy for the area to be improved, and output soil improvement suggestions.

10. The improved system based on salinity analysis of saline-alkali land according to claim 9, characterized in that: Also includes: Soil salinity prediction module: Based on the historical data collected at the sampling point, which consists of environmental data and soil moisture data, and using the data collected in the previous N days, predict the soil conductivity EC and soil pH value that characterize the soil salinity at the sampling point in the next N days. Combined with the currently collected soil temperature, soil moisture and groundwater level as the soil temperature, soil moisture and groundwater level in the next n days, as well as the environmental data obtained by the weather forecast in the next n days, the online real-time monitoring model constructed in the online real-time monitoring module is used to predict the soil salinity of the entire target farming area in the next n days. The forecast result map, so that operators can predict the trend of soil salinization and guide improvement strategies; Update module: Based on the collected data at the sampling points, the online real-time monitoring model is updated regularly, and the soil salinity prediction model is continuously updated to improve the model prediction accuracy, so as to more accurately match the improvement strategy and ensure the improvement effect.

Images