A method for co-regulating water and salt in saline-alkali soil based on multi-parameter sensing of internet of things

Abstract

The application discloses a kind of based on thing networking multi-parameter perception's saline-alkali soil water salt synergic regulation method, belong to saline-alkali soil improvement technical field.The steps of the present application are as follows: step one, collect the crop stem acoustic emission signal and soil water salt parameter of saline-alkali soil;Step two, crop stem acoustic emission signal is pretreated;Step three, feature is extracted based on xylem plug effective event;Step four, generate classification result and confidence;Step five, soil water salt parameter and crop stem acoustic emission signal are spatiotemporal alignment;Step six, the classification result of random forest classification model is verified;Step seven, stress index is calculated based on the extracted xylem plug effective event rate;Step eight, the classification result after verification and stress grade generate irrigation scheme based on preset irrigation mapping rule;Step nine, irrigation is carried out based on the generated irrigation scheme.The present application can reduce the probability that crop root suffers irreversible damage, improve crop growth effect.

Description

Technical Field

[0001] This invention relates to the field of saline-alkali land improvement technology, specifically to a method for coordinated regulation of water and salt in saline-alkali land based on multi-parameter sensing via the Internet of Things. Background Technology

[0002] Water-salt synergistic regulation of saline-alkali land refers to the scientific management of irrigation and drainage to control soil salinity within a range that crops can tolerate while meeting their water needs for growth, thereby achieving a dual balance of water and salt.

[0003] Current methods for coordinated water and salt regulation in saline-alkali land generally involve detecting water and salt stress in the soil using soil sensors before implementing coordinated water and salt regulation. By this time, the crop root system has already suffered irreversible damage to some extent, resulting in a response lag problem. Summary of the Invention

[0004] The purpose of this invention is to provide a method for coordinated water and salt regulation in saline-alkali land based on multi-parameter sensing via the Internet of Things. This method is characterized by its ability to construct crop xylem embolism based on acoustic emission signals, predict stress conditions in advance based on crop xylem embolism, and promptly and accurately regulate water and salt in a coordinated manner. It has the advantages of timely response, reducing the probability of irreversible damage to crop roots, and improving crop growth, thereby solving the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for coordinated regulation of water and salt in saline-alkali land based on multi-parameter sensing via the Internet of Things, comprising the following steps:

[0006] Step 1: Collect acoustic emission signals from crop stems and soil water and salt parameters in saline-alkali land. Soil water and salt parameters include soil volumetric water content and soil electrical conductivity.

[0007] Step 2: Preprocess the acoustic emission signal of crop stems, including bandpass filtering for noise reduction and extraction of effective xylem plugging events;

[0008] Step 3: Extract features based on effective xylem embolism events, including amplitude, duration, integral energy, rise time, dominant frequency, and event rate;

[0009] Step 4: The constructed random forest classification model generates classification results and confidence scores based on the extracted effective event features of xylem embolism;

[0010] Step 5: Spatiotemporally align soil water and salt parameters with the acoustic emission signals from crop stems;

[0011] Step 6: Verify the classification results of the random forest classification model using spatiotemporally aligned soil water and salt parameters;

[0012] Step 7: Calculate the stress index based on the extracted effective xylem embolism event rate. The stress index generates a stress level based on a preset threshold and a stress level mapping rule.

[0013] Step 8: Based on the verified classification results and stress levels, an irrigation plan is generated according to the preset irrigation mapping rules;

[0014] Step 9: Irrigate according to the generated irrigation plan to complete the intelligent coordinated regulation of water and salt in saline-alkali land based on IoT multi-parameter sensing.

[0015] Furthermore, in step two, the specific steps for extracting the effective events of xylem embolism in the acoustic emission signal of the crop stem are as follows:

[0016] The root mean square value of background noise is calculated based on the first second of the pulseless background segment of the acoustic emission signal of crop stalks in saline-alkali land after bandpass filtering and denoising.

[0017] The threshold for triggering xylem embolism events is preset based on the root mean square value;

[0018] Candidate valid xylem embolism events were selected from acoustic emission signals of crop stems in saline-alkali land based on xylem embolism event triggering thresholds.

[0019] Based on a preset threshold, valid xylem embolism events are filtered to extract them.

[0020] Furthermore, in step four, the constructed random forest classification model includes a data preprocessing module, a decision tree-based learner cluster, a voting module, and a confidence calculation module;

[0021] The data preprocessing module receives the extracted valid event features of xylem embolism and performs normalization processing to generate standardized feature vectors;

[0022] The decision tree-based learner cluster consists of several parallel independent CART classification decision trees. Each CART classification decision tree receives a standardized feature vector and performs classification reasoning based on the learned feature splitting rules to generate the corresponding classification result.

[0023] The voting module counts the classification results generated by all CART classification decision trees and uses the category with the most votes as the final classification result.

[0024] The confidence calculation module calculates the confidence score based on the statistical count of the final classification results.

[0025] Furthermore, in step five, the spatiotemporal alignment of the soil water and salt parameters with the acoustic emission signal of the crop stem includes:

[0026] Time alignment of acoustic emission signals from crop stems in saline-alkali land and soil water and salt parameters in saline-alkali land is based on timestamps;

[0027] The spatial alignment of acoustic emission signals from crop stems in saline-alkali land and soil water and salt parameters in saline-alkali land is based on the same coordinate system.

[0028] The acoustic emission signals of crop stems in saline-alkali land and the water and salt parameters of saline-alkali land soil are transformed to the same spatial reference coordinate system based on the calibrated coordinate transformation matrix;

[0029]

[0030] In the formula: This represents the coordinates of acoustic emission signals from crop stems in saline-alkali land or the soil water and salt parameters in saline-alkali land. Indicates the calibration rotation matrix; Indicates the calibration translation vector; Represents coordinates within the same spatial reference coordinate system;

[0031] Spatial alignment is performed based on coordinates from the same spatial reference coordinate system.

[0032] Furthermore, in step six, the verification step of the soil water and salt parameters on the classification results of the random forest classification model includes:

[0033] Soil volumetric water content and soil electrical conductivity are maintained or corrected based on comparison with preset soil volumetric water content and soil electrical conductivity thresholds.

[0034] Furthermore, in step seven, the expression for the stress index is:

[0035]

[0036] In the formula: This represents the acoustic emission event rate of current effective xylem embolism events; This indicates the calibrated acoustic emission event rate under normal conditions. This represents the maximum acoustic emission event rate.

[0037] Furthermore, in step eight, the preset irrigation mapping rules include:

[0038] If the classification result is normal, trigger the basic irrigation command to maintain basic irrigation;

[0039] If the classification result indicates a water stress state, trigger a water replenishment irrigation command to maintain the water replenishment irrigation amount;

[0040] Water replenishment irrigation amount:

[0041]

[0042] In the formula: Indicates the coercion level enhancement coefficient; Indicates field holding capacity; This indicates the current volumetric water content of the soil; Indicates the known root depth of a crop; Indicates the area of ​​the irrigated region;

[0043] If the classification result indicates a salt stress state, trigger the salt leaching irrigation command to maintain the leaching irrigation rate;

[0044] Leaching irrigation volume:

[0045]

[0046] In the formula: Indicates the coercion level enhancement coefficient; Indicates the electrical conductivity of irrigation water; Indicates the critical value for crop salt tolerance; Represents crop evapotranspiration;

[0047] If the classification result is a compound stress state, trigger the pulse irrigation command and maintain the pulse irrigation amount;

[0048] Pulse irrigation volume:

[0049]

[0050] In the formula: Indicates the coercion level enhancement coefficient; Indicates the pulse irrigation frequency; Indicates the amount of water used for a single pulse irrigation; Indicates the duration of a single pulse.

[0051] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0052] This invention periodically acquires acoustic emission signals from crop stems and soil water and salt parameters in saline-alkali land. It extracts features of effective xylem embolism events from the acoustic emission signals and constructs a random forest classification model to classify crop stress states based on these features. The soil water and salt parameters are used for secondary verification of the classification results. The classification results and the stress levels calculated from the acoustic emission signal features are used to generate irrigation plans based on preset irrigation mapping rules. Irrigation is then carried out based on these plans to achieve coordinated water and salt regulation in saline-alkali land. Compared to existing technologies, this invention can construct crop xylem embolism status based on acoustic emission signals, predict stress conditions in advance based on xylem embolism status, and implement timely and accurate coordinated water and salt regulation. This timely response reduces the probability of irreversible damage to crop roots and improves crop growth. Attached Figure Description

[0053] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0054] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0055] Example 1

[0056] Please see Figure 1 A method for coordinated regulation of water and salt in saline-alkali land based on multi-parameter sensing via the Internet of Things includes the following steps:

[0057] Acoustic emission signals from crop stems in saline-alkali land were collected.

[0058] Preprocessing was performed on the collected acoustic emission signals from crop stems in saline-alkali land, including bandpass filtering for noise reduction and extraction of effective xylem embolism events;

[0059] Extraction of valid events related to xylem embolism in acoustic emission signals from crop stems in saline-alkali soil:

[0060] The root mean square value of background noise is calculated based on the first second of the pulseless background segment of the acoustic emission signal of crop stalks in saline-alkali land after bandpass filtering and denoising.

[0061] Root mean square value:

[0062]

[0063] In the formula: Indicates the amplitude of the sampling points in the background segment; Indicates the number of sampling points;

[0064] The threshold for triggering xylem embolism events is preset based on the root mean square value;

[0065] Candidate valid xylem embolism events were selected from acoustic emission signals of crop stems in saline-alkali land based on xylem embolism event triggering thresholds.

[0066] A 30-second continuous signal of acoustic emission from crop stems in saline-alkali land after bandpass filtering and denoising is scanned point by point. When the amplitude of three consecutive sampling points exceeds the xylem embolism event trigger threshold, it is determined to be the starting point of the xylem embolism event. The point with the largest amplitude is recorded from the starting point as the peak point. When the amplitude of five consecutive sampling points falls back to below the xylem embolism event trigger threshold, it is determined to be the ending point of the xylem embolism event. The signal segment from the starting point to the ending point of the xylem embolism event is extracted as a candidate valid xylem embolism event.

[0067] Based on a preset threshold filtering of candidate xylem embolism valid events, valid xylem embolism events are extracted.

[0068] Candidate xylem embolism effective event filtering:

[0069] Candidate xylem embolism effective events with a duration shorter than the preset time were removed. Most of these candidate xylem embolism effective events originated from transient interferences such as field electromagnetic interference and wind noise.

[0070] Candidate xylem embolism effective events with peak values ​​smaller than the preset peak value amplitude are removed, as most of these candidate xylem embolism effective events originate from background noise.

[0071] Eliminate candidate xylem embolism events whose integral energy is less than the preset event integral energy;

[0072] Event Integral Energy:

[0073] The voltage integral is calculated by summing the voltages at all sampling points of the candidate xylem embolism valid event.

[0074]

[0075] In the formula: Represents the voltage integral; Indicates the voltage at the sampling point; Indicates the sampling interval; This represents the total number of valid sampling points for candidate xylem embolism events;

[0076] Voltage integration is converted into integrated energy based on a calibrated conversion factor;

[0077]

[0078] In the formula: Represents integral energy; Indicates the calibration conversion factor; Represents the voltage integral;

[0079] Based on the effective event extraction features of xylem embolism, including amplitude, duration, integral energy, rise time, dominant frequency and event rate;

[0080] Amplitude: Converts the peak amplitude of an effective xylem embolism event into sound pressure level;

[0081]

[0082] In the formula: Indicates the peak amplitude of effective xylem embolism events; Indicates the calibrated sensitivity; Indicates the standard reference sound pressure level;

[0083] The average sound pressure level of all effective xylem embolism events was calculated as an amplitude characteristic to characterize the degree of stress.

[0084] Duration: The time from the start of the effective xylem embolism event to the end of the effective xylem embolism event, characterizing the severity of xylem embolism;

[0085] Integral energy: The voltage integral is calculated by summing the voltages at all sampling points of an effective xylem embolism event, and converted into integral energy based on the calibrated equipment conversion coefficient, which characterizes the stress level;

[0086] Rise time: The time from the start of the effective xylem embolism event to the peak point, characterizing the rate of embolism formation;

[0087] Dominant frequency: Based on the effective event waveform of xylem embolism, a fast Fourier transform is performed to convert the time domain signal into a frequency domain signal. The frequency point with the maximum amplitude in the frequency domain signal is found as the dominant frequency to distinguish xylem embolism from environmental noise.

[0088] Event rate: The number of effective xylem embolism events within a preset time period, representing the degree of stress;

[0089] A random forest classification model is constructed, and classification results and confidence scores are generated based on the extracted effective event features of xylem embolism.

[0090] The random forest classification model includes a data preprocessing module, a decision tree-based learner cluster, a voting module, and a confidence calculation module;

[0091] The data preprocessing module receives the extracted valid event features of xylem embolism and performs normalization processing to generate standardized feature vectors;

[0092] The decision tree-based learner cluster consists of several parallel independent CART classification decision trees. Each CART classification decision tree receives a standardized feature vector and performs classification reasoning based on the learned feature splitting rules to generate the corresponding classification result.

[0093] The voting module counts the classification results generated by all CART classification decision trees and uses the category with the most votes as the final classification result.

[0094] The confidence calculation module calculates the confidence level based on the statistical count of the final classification results;

[0095] Confidence level:

[0096]

[0097] In the formula: This indicates that the current sample is classified as a category. The number of CART classification decision trees; This represents the total number of CART classification decision trees in the random forest;

[0098] Collect historical acoustic emission signals from crop stems in saline-alkali land;

[0099] Preprocessing was performed on the historical acoustic emission signals of crop stems collected in saline-alkali land, including bandpass filtering for noise reduction and extraction of effective xylem embolism events;

[0100] Based on the effective event extraction features of xylem embolism, including amplitude, duration, integral energy, rise time, dominant frequency and event rate;

[0101] Stress type labels that characterize effective xylem embolism events include normal state, moisture stress state, salt stress state, and combined stress state;

[0102] An acoustic emission feature dataset is constructed based on the effective event characteristics of xylem embolism and the matched stress type labels;

[0103] The acoustic emission feature dataset was divided into training, validation and test sets in a 7:2:1 ratio.

[0104] The random forest classification model is trained based on the training set. During the training process, random sampling with replacement is performed from the training set using the bootstrap sampling method to generate a training subset for each CART classification decision tree. During the node splitting process of each CART classification decision tree, some features in the training subset are randomly selected and splitting training is performed using the Gini coefficient as the node splitting criterion until the number of node samples is less than the threshold. This completes the construction of each CART classification decision tree. The above process is repeated to generate a preset number of CART classification decision trees, forming a decision tree base learner cluster.

[0105] Optimize the hyperparameters of a random forest classification model based on the validation set;

[0106] The classification accuracy of the random forest classification model is calculated based on the test set, and the random forest classification model is evaluated based on the classification accuracy until the optimal result is achieved.

[0107] Soil water and salt parameters of saline-alkali land were collected, including soil volumetric water content and soil electrical conductivity.

[0108] Spatiotemporal alignment was performed based on the collected acoustic emission signals of crop stems in saline-alkali land and the water and salt parameters of saline-alkali land soil.

[0109] Time alignment of acoustic emission signals from crop stems in saline-alkali land and soil water and salt parameters in saline-alkali land is based on timestamps;

[0110] The spatial alignment of acoustic emission signals from crop stems in saline-alkali land and soil water and salt parameters in saline-alkali land is based on the same coordinate system.

[0111] The acoustic emission signals of crop stems in saline-alkali land and the water and salt parameters of saline-alkali land soil are transformed to the same spatial reference coordinate system based on the calibrated coordinate transformation matrix;

[0112]

[0113] In the formula: This represents the coordinates of acoustic emission signals from crop stems in saline-alkali land or the soil water and salt parameters in saline-alkali land. Indicates the calibration rotation matrix; Indicates the calibration translation vector; Represents coordinates within the same spatial reference coordinate system;

[0114] Spatial alignment based on coordinates from the same spatial reference coordinate system;

[0115] The classification results were validated based on the generated classification results and the spatiotemporally aligned water and salt parameters of saline-alkali land.

[0116] If the classification result is normal, then it is determined whether the soil volumetric water content is higher than the preset soil volumetric water content and whether the soil electrical conductivity is lower than the preset soil electrical conductivity. If so, it is determined to be normal. Otherwise, if the soil volumetric water content is lower than the preset soil volumetric water content, it is modified to water stress state. If the soil electrical conductivity is higher than the preset soil electrical conductivity, it is modified to salt stress state. If neither of them is met, it is modified to composite stress state and fed back to the random forest classification model for iterative optimization.

[0117] If the classification result is a water stress state, then it is determined whether the soil volumetric water content is lower than the preset soil volumetric water content. If the soil volumetric water content is lower than the preset soil volumetric water content, it is determined to be a water stress state; otherwise, it is modified to a normal state and fed back to the random forest classification model for iterative optimization.

[0118] If the classification result is salt stress, then it is determined whether the soil electrical conductivity is higher than the preset soil electrical conductivity. If so, it is determined to be salt stress; otherwise, it is modified to normal state and fed back to the random forest classification model for iterative optimization.

[0119] If the classification result is a combined stress state, check whether the soil volumetric water content is lower than the preset soil volumetric water content and whether the soil electrical conductivity is higher than the preset soil electrical conductivity. If yes, it is determined to be a combined stress state. Otherwise, if the soil volumetric water content is higher than the preset soil volumetric water content, it is changed to a salt stress state. If the soil electrical conductivity is lower than the preset soil electrical conductivity, it is changed to a water stress state. If neither of them is met, it is changed to a normal state and fed back to the random forest classification model for iterative optimization.

[0120] The stress index is calculated based on the event rate characteristics of the extracted effective xylem embolism events;

[0121] Stress Index:

[0122]

[0123] In the formula: This represents the acoustic emission event rate of current effective xylem embolism events; This indicates the calibrated acoustic emission event rate under normal conditions. Indicates the maximum acoustic emission event rate;

[0124] The stress index generates stress levels based on preset thresholds and stress level mapping rules;

[0125] The classification results and stress levels generate irrigation plans based on preset irrigation mapping rules;

[0126] If the classification result is normal, trigger the basic irrigation command to maintain basic irrigation;

[0127] If the classification result indicates a water stress state, trigger a water replenishment irrigation command to maintain the water replenishment irrigation amount;

[0128] Water replenishment irrigation amount:

[0129]

[0130] In the formula: Indicates the coercion level enhancement coefficient; Indicates field holding capacity; This indicates the current volumetric water content of the soil; Indicates the known root depth of a crop; Indicates the area of ​​the irrigated region;

[0131] If the classification result indicates a salt stress state, trigger the salt leaching irrigation command to maintain the leaching irrigation rate;

[0132] Leaching irrigation volume:

[0133]

[0134] In the formula: Indicates the coercion level enhancement coefficient; Indicates the electrical conductivity of irrigation water; Indicates the critical value for crop salt tolerance; Represents crop evapotranspiration;

[0135] If the classification result is a compound stress state, trigger the pulse irrigation command and maintain the pulse irrigation amount;

[0136] Pulse irrigation volume:

[0137]

[0138] In the formula: Indicates the coercion level enhancement coefficient; Indicates the pulse irrigation frequency; Indicates the amount of water used for a single pulse irrigation; Indicates the duration of a single pulse;

[0139] Irrigation is carried out based on the generated irrigation plan, and intelligent coordinated regulation of water and salt in saline-alkali land is achieved based on multi-parameter sensing of the Internet of Things.

[0140] Sunflowers were planted in a saline-alkali experimental field using conventional planting methods, and water and salt regulation was carried out based on the above methods:

[0141] The classification is based on a random forest classification model, including normal state, water stress state, salt stress state and combined stress state;

[0142] The stress index is calculated based on the event rate characteristics of the extracted effective xylem embolism events;

[0143] The classification results and stress levels generate irrigation plans based on preset irrigation mapping rules;

[0144] If the classification result is normal, trigger the basic irrigation command to maintain basic irrigation;

[0145] If the classification result indicates a water stress state, trigger a water replenishment irrigation command to maintain the water replenishment irrigation amount;

[0146] If the classification result indicates a salt stress state, trigger the salt leaching irrigation command to maintain the leaching irrigation rate;

[0147] If the classification result is a compound stress state, trigger the pulse irrigation command and maintain the pulse irrigation amount;

[0148] Irrigation is carried out based on the generated irrigation plan, and intelligent coordinated regulation of water and salt in saline-alkali land is achieved based on multi-parameter sensing of the Internet of Things.

[0149] Comparative Example 1

[0150] Sunflowers were planted in a saline-alkali experimental field using conventional planting methods, and water and salt regulation was performed based on soil sensor data.

[0151] Soil water and salt parameters of saline-alkali land are collected based on soil sensors, including soil volumetric water content and soil electrical conductivity.

[0152] The state of saline-alkali land is determined based on soil volumetric water content and soil electrical conductivity, including normal state, water stress state, salt stress state and combined stress state.

[0153] State judgment: Determine whether the soil volumetric water content is higher than the preset soil volumetric water content and whether the soil electrical conductivity is lower than the preset soil electrical conductivity. If so, it is judged as a normal state. Conversely, if the soil volumetric water content is lower than the preset soil volumetric water content, it is judged as a water stress state. If the soil electrical conductivity is higher than the preset soil electrical conductivity, it is judged as a salt stress state. If neither of the above conditions are met, it is judged as a combined stress state.

[0154] Irrigation schemes are generated based on the state of saline-alkali land:

[0155] If the situation is normal, trigger the basic irrigation command to maintain basic irrigation;

[0156] If the situation is under water stress, trigger the water replenishment irrigation command to maintain the water replenishment irrigation amount;

[0157] If the situation is under salt stress, trigger the salt leaching irrigation command to maintain the leaching irrigation rate;

[0158] If the situation is under combined stress, trigger the pulse irrigation command and maintain the pulse irrigation amount;

[0159] Irrigation is carried out based on the generated irrigation plan, and intelligent coordinated regulation of water and salt in saline-alkali land is achieved based on soil parameter perception.

[0160] Comparative Example 2

[0161] Sunflowers were planted in a saline-alkali experimental field using conventional planting methods, with natural irrigation based on experience.

[0162] Based on experience, the state of saline-alkali land is judged, including normal state, water stress state, salt stress state and combined stress state;

[0163] Irrigation schemes are generated based on the state of saline-alkali land:

[0164] If the situation is normal, trigger the basic irrigation command to maintain basic irrigation;

[0165] If the situation is under water stress, trigger the water replenishment irrigation command to maintain the water replenishment irrigation amount;

[0166] If the situation is under salt stress, trigger the salt leaching irrigation command to maintain the leaching irrigation rate;

[0167] If the situation is under combined stress, trigger the pulse irrigation command and maintain the pulse irrigation amount;

[0168] Irrigation is carried out based on the generated irrigation plan, and intelligent coordinated regulation of water and salt in saline-alkali land is achieved based on soil parameter perception.

[0169] The results of soil water and salt dynamics, crop growth data, and yield and quality data for Example 1, Comparative Example 1, and Comparative Example 2 are as follows:

[0170] Changes in soil salinity in the 0-20cm layer:

[0171] deal with Pre-sowing salt content (%) Seedling stage (%) Flowering period (%) Harvest period (%) Desalination rate (%) Example 1 0.55 0.48 0.42 0.38 -31 Comparative Example 1 0.55 0.52 0.48 0.45 -18 Comparative Example 2 0.55 0.58 0.62 0.68 +24

[0172] Salinity changes in soil layers 20-40cm:

[0173] deal with Pre-sowing salt content (%) Harvest period (%) Salt displacement Example 1 0.48 0.55 +0.07 Comparative Example 1 0.48 0.52 +0.04 Comparative Example 2 0.48 0.45 -0.03

[0174] Plant height dynamics:

[0175] deal with End of seedling stage (cm) Budding stage (cm) Flowering period (cm) Grouting period (cm) Example 1 45±3 128±5 168±4 172±4 Comparative Example 1 44±3 122±6 160±6 165±6 Comparative Example 2 42±4 115±8 152±10 158±9

[0176] Yield:

[0177] deal with Disc diameter (cm) Number of discs 1000-grain weight (g) Grain yield (kg / ha) Example 1 21.5±1.2 1120±80 6.8±0.3 2980±135 Comparative Example 1 19.8±1.0 1050±90 6.5±0.3 2680±145 Comparative Example 2 18.2±1.5 950±120 6.2±0.4 2400±180

[0178] quality:

[0179] deal with Oil content (%) Protein content (%) Oleic acid content (%) Product grade Example 1 44.2±0.8 18.5±0.5 32.5±1.2 Level 1 Comparative Example 1 41.8±0.9 19.2±0.6 30.1±1.3 Level 1 Comparative Example 2 39.5±1.2 20.2±0.8 28.3±1.5 Level 2

[0180] As can be seen from the table above, the water-salt synergistic regulation method for saline-alkali land of the present invention has the best effect.

[0181] In summary, this invention periodically acquires acoustic emission signals from crop stems and soil water and salt parameters in saline-alkali land, extracts features of effective xylem embolism events from the acoustic emission signals, and constructs a random forest classification model to classify crop stress states based on the features of effective xylem embolism events. Soil water and salt parameters are used for secondary verification of the classification results. The classification results and the stress levels calculated from the acoustic emission signal features are used to generate irrigation schemes based on preset irrigation mapping rules. Irrigation is then carried out based on these schemes to achieve coordinated water and salt regulation in saline-alkali land. Compared with existing technologies, this invention can construct crop xylem embolism status based on acoustic emission signals, predict stress conditions in advance based on the crop xylem embolism status, and promptly and accurately regulate water and salt. The timely response reduces the probability of irreversible damage to crop roots and improves crop growth.

[0182] The present invention and its embodiments have been described above illustratively. This description is not restrictive, and the figures shown are only one embodiment of the present invention; the actual structure is not limited thereto. Therefore, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the present invention, such designs should fall within the protection scope of the present invention.

Claims

1. A method for coordinated regulation of water and salt in saline-alkali land based on multi-parameter sensing via the Internet of Things, characterized in that: Includes the following steps: Step 1: Collect acoustic emission signals from crop stems and soil water and salt parameters in saline-alkali land. Soil water and salt parameters include soil volumetric water content and soil electrical conductivity. Step 2: Preprocess the acoustic emission signal of crop stems, including bandpass filtering for noise reduction and extraction of effective xylem plugging events; Step 3: Extract features based on effective xylem embolism events, including amplitude, duration, integral energy, rise time, dominant frequency, and event rate; Step 4: The constructed random forest classification model generates classification results and confidence scores based on the extracted effective event features of xylem embolism; Step 5: Spatiotemporally align soil water and salt parameters with the acoustic emission signals from crop stems; Step 6: Verify the classification results of the random forest classification model using spatiotemporally aligned soil water and salt parameters; Step 7: Calculate the stress index based on the extracted effective xylem embolism event rate. The stress index generates a stress level based on a preset threshold and a stress level mapping rule. Step 8: Based on the verified classification results and stress levels, an irrigation plan is generated according to the preset irrigation mapping rules; Step 9: Irrigate according to the generated irrigation plan to complete the intelligent coordinated regulation of water and salt in saline-alkali land based on IoT multi-parameter sensing.

2. The method for coordinated regulation of water and salt in saline-alkali land based on multi-parameter sensing via the Internet of Things, as described in claim 1, is characterized in that: In step two, the specific steps for extracting the effective events of xylem embolism in the acoustic emission signal of the crop stem are as follows: The root mean square value of background noise is calculated based on the first second of the pulseless background segment of the acoustic emission signal of crop stalks in saline-alkali land after bandpass filtering and denoising. The threshold for triggering xylem embolism events is preset based on the root mean square value; Candidate valid xylem embolism events were selected from acoustic emission signals of crop stems in saline-alkali land based on xylem embolism event triggering thresholds. Based on a preset threshold, valid xylem embolism events are filtered to extract them.

3. The method for coordinated regulation of water and salt in saline-alkali land based on multi-parameter sensing via the Internet of Things, as described in claim 1, is characterized in that: In step four, the constructed random forest classification model includes a data preprocessing module, a decision tree-based learner cluster, a voting module, and a confidence calculation module. The data preprocessing module receives the extracted valid event features of xylem embolism and performs normalization processing to generate standardized feature vectors; The decision tree-based learner cluster consists of several parallel independent CART classification decision trees. Each CART classification decision tree receives a standardized feature vector and performs classification reasoning based on the learned feature splitting rules to generate the corresponding classification result. The voting module counts the classification results generated by all CART classification decision trees and uses the category with the most votes as the final classification result. The confidence calculation module calculates the confidence score based on the statistical count of the final classification results.

4. The method for coordinated regulation of water and salt in saline-alkali land based on multi-parameter sensing via the Internet of Things, as described in claim 1, is characterized in that: In step five, the spatiotemporal alignment of the soil water and salt parameters with the acoustic emission signal of the crop stem includes: Time alignment of acoustic emission signals from crop stems in saline-alkali land and soil water and salt parameters in saline-alkali land is based on timestamps; The spatial alignment of acoustic emission signals from crop stems in saline-alkali land and soil water and salt parameters in saline-alkali land is based on the same coordinate system. The acoustic emission signals of crop stems in saline-alkali land and the water and salt parameters of saline-alkali land soil are transformed to the same spatial reference coordinate system based on the calibrated coordinate transformation matrix; In the formula: This represents the coordinates of acoustic emission signals from crop stems in saline-alkali land or the soil water and salt parameters in saline-alkali land. Indicates the calibration rotation matrix; Indicates the calibration translation vector; Represents coordinates within the same spatial reference coordinate system; Spatial alignment is performed based on coordinates from the same spatial reference coordinate system.

5. The method for coordinated regulation of water and salt in saline-alkali land based on multi-parameter sensing via the Internet of Things, as described in claim 1, is characterized in that: Step six, the verification steps for the soil water and salt parameters on the classification results of the random forest classification model, include: Soil volumetric water content and soil electrical conductivity are maintained or corrected based on comparison with preset soil volumetric water content and soil electrical conductivity thresholds.

6. The method for coordinated regulation of water and salt in saline-alkali land based on multi-parameter sensing via the Internet of Things, as described in claim 1, is characterized in that: In step seven, the expression for the stress index is: In the formula: This represents the acoustic emission event rate of current effective xylem embolism events; This indicates the calibrated acoustic emission event rate under normal conditions. This represents the maximum acoustic emission event rate.

7. The method for coordinated regulation of water and salt in saline-alkali land based on multi-parameter sensing via the Internet of Things, as described in claim 1, is characterized in that: In step eight, the preset irrigation mapping rules include: If the classification result is normal, trigger the basic irrigation command to maintain basic irrigation; If the classification result indicates a water stress state, trigger a water replenishment irrigation command to maintain the water replenishment irrigation amount; Water replenishment irrigation amount: In the formula: Indicates the coercion level enhancement coefficient; Indicates field holding capacity; This indicates the current volumetric water content of the soil; Indicates the known root depth of a crop; Indicates the area of ​​the irrigated region; If the classification result indicates a salt stress state, trigger the salt leaching irrigation command to maintain the leaching irrigation rate; Leaching irrigation volume: In the formula: Indicates the coercion level enhancement coefficient; Indicates the electrical conductivity of irrigation water; Indicates the critical value for crop salt tolerance; Represents crop evapotranspiration; If the classification result is a compound stress state, trigger the pulse irrigation command and maintain the pulse irrigation amount; Pulse irrigation volume: In the formula: Indicates the coercion level enhancement coefficient; Indicates the pulse irrigation frequency; Indicates the amount of water used for a single pulse irrigation; Indicates the duration of a single pulse.

Images