Desert area water resource intelligent deployment system and method based on internet of things

Abstract

The application discloses a kind of desert area water resource intelligent deployment system and method based on Internet of Things.The method comprises the following steps: dividing desert area into several management areas, obtaining soil data, vegetation species and vegetation physiological parameters in each management area, and available water resource data of each water source point in desert area;Based on the vegetation species and vegetation physiological parameters in each management area, determine the vegetation water requirement estimation data of each management area;Using the soil data, the vegetation water requirement estimation data of each management area is corrected to generate comprehensive water resource demand data of each management area;Based on the comprehensive water resource demand data of each management area and the available water resource data of each water source point, generate the water resource scheduling scheme of desert area.The application effectively improves the problem of low-efficiency configuration of desert water resources.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of water resource allocation, and more particularly, to a desert area water resource intelligent allocation system and method based on the Internet of Things. BACKGROUND

[0002] Desert areas account for more than 40% of the global land area and continue to expand with climate change. Current desert area management is mostly through vegetation sand fixation and local ecological environment restoration to inhibit the expansion of desert area. However, ecological restoration relies on efficient water resource allocation. Due to the extreme scarcity of water resources in desertification areas and uneven spatiotemporal distribution, existing water resource scheduling schemes or technologies have multidimensional defects, which cannot better meet the water demand of different regions in the desert area.

[0003] Specifically, the desert area water resource allocation has long relied on artificial experience or static models, and has the following problems. The existing method only focuses on single soil moisture or meteorological data, without fusing key parameters such as vegetation physiological state (such as stomatal conductance, vegetation water potential), soil salinization index (conductivity), and temperature, resulting in distorted water demand prediction, i.e., the intensified vegetation transpiration under high temperature stress is not quantified, and the salt accumulation inhibiting root water absorption is not corrected. The traditional scheme relies on the static lookup table method of vegetation type and growth stage, ignoring the dynamic influence of real-time physiological state (health / stress level) of vegetation on water demand, and cannot respond to the change of water demand law caused by drought or salt stress. The existing technology completely ignores the soil salinization management demand. When the soil conductivity exceeds the salt tolerance threshold of vegetation, there is no leaching water demand calculation mechanism, resulting in continuous accumulation of salt. The existing water resource scheduling model is not adapted to the multi-source data architecture of the Internet of Things, neither compatible with real-time heterogeneous data of soil-vegetation-weather-water source, nor optimized in coordination of water supply capacity, water source supply limit and salinization leaching demand, ultimately leading to unreasonable water resource allocation, low efficiency of water use, and poor ecological restoration effect. SUMMARY

[0004] The main purpose of the present application is to provide a desert area water resource intelligent allocation system and method based on the Internet of Things, to solve the problems of unreasonable water resource allocation, low efficiency of water use, and poor ecological restoration effect in the prior art.

[0005] To achieve the above purpose, the present application provides the following technical scheme:

[0006] One aspect of the present application provides a desert area water resource intelligent allocation method based on the Internet of Things, comprising:

[0007] dividing the desert area into a plurality of management regions, obtaining soil data, vegetation species and vegetation physiological parameters in each management region, and available water resource data of each water source point in the desert area;

[0008] determine vegetation water requirement estimation data of each management area based on the vegetation types and the vegetation physiological parameters in each management area;

[0009] correct the vegetation water requirement estimation data of each management area by using the soil data, and generate comprehensive water resource demand data of each management area;

[0010] generate a water resource scheduling scheme for the desert area based on the comprehensive water resource demand data of each management area and the available water resource data of each water source point.

[0011] In one embodiment, the soil data includes soil type, soil conductivity, soil water content, etc. The vegetation physiological parameters include leaf water content, stomatal conductance, transpiration rate, photosynthetic rate, chlorophyll content, vegetation water potential, and environmental temperature, etc.

[0012] In one embodiment, the vegetation water requirement estimation data of each management area is obtained in the following manner:

[0013] statistically determine the distribution proportion of different vegetation types in each management area, and obtain the growth stages of different vegetation types;

[0014] match corresponding basic water requirement data from a preset vegetation water requirement characteristic database according to the distribution proportion of vegetation types and the growth stages of different vegetation types, so as to determine the basic water requirement data of each management area;

[0015] collect the vegetation physiological parameters of different vegetation in each management area, and perform time series data analysis on the collected vegetation physiological parameters to identify the physiological state of the vegetation;

[0016] adjust the basic water requirement data of each management area according to the physiological state of the vegetation, to obtain the vegetation water requirement estimation data of each management area.

[0017] In one embodiment, the growth stages of different vegetation types are obtained in the following manner:

[0018] collect images of each vegetation type at preset time intervals, and pre-process the collected vegetation images to form a time series image data set bound to each vegetation type;

[0019] extract features from the images in the time series image data set, and generate appearance feature vector time series of each vegetation type;

[0020] input the appearance feature vector time series into a ResNet-50 convolutional neural network with removed full connection layer, to obtain a time series feature vector representing the growth dynamics of the vegetation;

[0021] Obtaining a feature template vector of different growth stages of the vegetation, and performing similarity matching between the time-series feature vector and the feature template vector of different growth stages of the vegetation, so as to determine the current growth stage of the vegetation.

[0022] In one embodiment, the manner of performing time-series data analysis on the collected vegetation physiological parameters and identifying the physiological state of the vegetation comprises:

[0023] Timestamp alignment, denoising, outlier rejection and normalization processing are performed on the collected physiological parameter data to form standardized vegetation physiological parameters, and a water stress index of the vegetation is calculated based on the standardized vegetation physiological parameters;

[0024] Synergistic analysis is performed on leaf water content, stomatal conductance, transpiration rate, photosynthetic rate, chlorophyll content and vegetation water potential to obtain the cooperation relationship between the vegetation physiological parameters; a feature index is constructed based on the cooperation relationship between the vegetation physiological parameters, and the constructed feature index is spliced to obtain a state feature vector;

[0025] Uniform manifold approximation and projection algorithm is used to perform dimension reduction processing on the state feature vector to obtain a resonance frequency space coordinate of the state feature vector; resonance intensity calculation is performed on the resonance frequency space coordinate to obtain a resonance intensity of the state feature vector;

[0026] Based on the water stress index and the resonance intensity, and in combination with preset classification thresholds of the water stress index and the resonance intensity, the physiological state of the vegetation is evaluated to obtain the physiological state of the vegetation.

[0027] In one embodiment, the manner of adjusting the basic water requirement data of each management area according to the physiological state of the vegetation comprises:

[0028] Based on the physiological state of the vegetation, the resonance intensity corresponding to the physiological state and the water stress index of the vegetation, a water requirement correction factor is generated in combination with the environmental temperature, and the water requirement correction factor is used to update the basic water requirement data of the vegetation, so as to obtain vegetation water requirement estimation data of each management area.

[0029] In one embodiment, the manner of correcting the vegetation water requirement estimation data of each management area by using the soil data comprises:

[0030] The minimum soil water content required for the growth of the vegetation in each management area and the salt tolerance threshold are obtained; if the soil electrical conductivity of the management area is greater than or equal to the electrical conductivity corresponding to the salt tolerance threshold of the vegetation in the corresponding management area, the available soil water resource data of the management area is zero; if the soil electrical conductivity of the management area is less than the electrical conductivity corresponding to the salt tolerance threshold of the vegetation in the corresponding management area, the available soil water resource data of the management area is calculated based on the difference between the soil water content and the minimum soil water content;

[0031] The soil water resource data of each management area is compared with the vegetation water demand estimation data of each management area to obtain water resource gap data of each management area;

[0032] The soil electrical conductivity threshold in each management area is set based on the salt tolerance threshold of the vegetation in each management area; based on the soil type, soil electrical conductivity and pre-set soil electrical conductivity threshold, the leaching water amount data of each management area is obtained through a pre-set leaching water demand calculation function;

[0033] The water resource gap data of each management area is added to the leaching water amount data to obtain comprehensive water resource demand data of each management area.

[0034] In one embodiment, the water resource scheduling scheme is generated in the following manner:

[0035] A water resource scheduling optimization model is constructed, and the optimization objective of the water resource scheduling optimization model is to minimize the water delivery cost, and the constraint condition is the available water resource supply capacity limit of each water source point and the maximum transmission capacity limit of each pipeline in the water delivery pipe network;

[0036] Based on the available water resource data of each water source point, the comprehensive water resource demand data of each management area, the optimization objective and the constraint condition, the water resource scheduling optimization model is solved by a network flow optimization algorithm to generate a water resource scheduling scheme.

[0037] Another aspect of the present application provides a desert area water resource intelligent allocation system based on Internet of Things, which is used to realize the desert area water resource intelligent allocation method based on Internet of Things, and the system comprises:

[0038] A data acquisition module is used to divide the desert area into a plurality of management areas, and obtain soil data, vegetation types, vegetation physiological parameters in each management area and available water resource data of each water source point in the desert area;

[0039] A water demand calculation module is used to determine the vegetation water demand estimation data of each management area based on the vegetation types and vegetation physiological parameters in each management area;

[0040] A water demand adjustment module is used to correct the vegetation water demand estimation data of each management area by using the soil data to generate comprehensive water resource demand data of each management area;

[0041] A scheme generation module is used to generate a water resource scheduling scheme of the desert area based on the comprehensive water resource demand data of each management area and the available water resource data of each water source point.

[0042] Compared with the prior art, the advantages of the present application at least lie in that: by dividing the desert grid management area and synchronously acquiring heterogeneous parameters such as soil conductivity, vegetation water potential, stomatal conductance and temperature, the precision and intelligence breakthrough of desert water resource allocation is realized: the innovative water demand prediction mechanism is based on dynamic hierarchical evaluation of vegetation physiological state (healthy / mild / moderate / severe stress), combined with real-time temperature data to generate a water demand correction factor, dynamically quantifying the water demand response of high temperature transpiration surge and salt drought combined stress through time series analysis of the coordinated change of physiological parameters, overcoming the response lag of traditional static models to the fluctuation of vegetation water use efficiency to reduce irrigation redundancy; the innovative salinization control path creates a leaching water demand calculation model for soil conductivity exceeding the salt tolerance threshold of vegetation, coupled with the water conductivity characteristics of soil types and the mechanics law of salt migration, accurately calculating the leaching water quantity and incorporating the demand when the conductivity sensor triggers a critical early warning, simultaneously achieving salt removal and invalid water consumption reduction to reverse the vegetation salt stress mortality; the global scheduling efficiency is transferred, a three-dimensional optimization model of water transmission pipe network transmission capacity constraint, water source supply limit constraint and leaching demand constraint is built, and scheduling instructions are generated with the minimum water transmission cost as the target, breaking through the limitations of single-point irrigation and water source island to compress the energy consumption and time delay of pipe network water transmission; the problem of inefficient allocation of desert water resources is effectively improved. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 FIG. 1 is a schematic diagram of a desert area water resource intelligent allocation method based on the Internet of Things according to an embodiment of the present application;

[0044] Figure 2 FIG. 2 is a schematic diagram of a desert area water resource intelligent allocation system based on the Internet of Things according to an embodiment of the present application. DETAILED DESCRIPTION

[0045] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor are within the scope of protection of the present application.

[0046] Embodiment 1

[0047] Please refer to Figure 1 The desert area water resource intelligent allocation method based on the Internet of Things provided by the present embodiment includes:

[0048] The desert area is divided into a plurality of management areas, and soil data, vegetation species, vegetation physiological parameters in each management area, and available water resource data of each water source point in the desert area are acquired;

[0049] Based on the vegetation types and vegetation physiological parameters in each management area, the vegetation water requirement estimation data of each management area is determined:

[0050] The vegetation water requirement estimation data of each management area is corrected by using the soil data, and comprehensive water resource demand data of each management area is generated;

[0051] Based on the comprehensive water resource demand data of each management area and the available water resource data of each water source point, a water resource scheduling scheme for the desert area is generated.

[0052] The soil data includes soil type (such as sandy soil, clay soil, loam soil), soil conductivity (reflecting the degree of salinization, the higher the degree of salinization, the stronger the conductivity) and soil water content; the vegetation physiological parameters include leaf water content, stomatal conductance, transpiration rate, photosynthetic rate, chlorophyll content and vegetation water potential;

[0053] Based on the soil type, superimposed with the river and lake water system distribution, using GIS spatial analysis technology and clustering algorithm, the desert area is divided into several groups of different ecological management areas; each group of management areas not only has highly similar soil physical and chemical properties, but also has relatively unified hydrological cycle process and ecological water demand law, forming management areas with spatial heterogeneity and ecological integrity;

[0054] In each management area, multiple Internet of Things sensors and monitoring equipment are deployed to collect soil conductivity (range 0-25 dS / m, accuracy ±0.1 dS / m) and soil water content; and through the deployment of multiple Internet of Things sensors, the physiological parameters of different vegetation types in each management area are obtained, and the average value of multiple data of the same parameter of the same vegetation type is taken as the final value; the vegetation physiological parameters include leaf water content (%), stomatal conductance (mol / m 2 s), transpiration rate (mmol / m 2 s), photosynthetic rate (μmol / m 2 s), chlorophyll content (SPAD unit), vegetation water potential (MPa) and environmental temperature (℃), etc., and are not limited thereto.

[0055] Further, the vegetation water requirement estimation data of each management area is obtained in the following manner:

[0056] The distribution proportion of different vegetation types in each management area is counted, and the growth stages of different vegetation in each management area are determined;

[0057] According to the distribution proportion of vegetation types and the growth stages of different vegetation types, the corresponding basic water requirement data is matched from the preset vegetation water requirement characteristic database, so as to determine the basic water requirement data of each management area;

[0058] Collecting the physiological parameters of different vegetation in each management area, and performing time series data analysis on the collected physiological parameters of the vegetation to identify the physiological state of the vegetation;

[0059] According to the physiological state of different vegetation, the basic water requirement data of each management area is adjusted to obtain the vegetation water requirement estimation data of each management area.

[0060] Specifically, a vegetation water requirement characteristic database is established, which includes the baseline water requirement of different vegetation in different growth stages (long leaf period, dormant period, flowering period, fruiting period, etc.) (such as unit biomass daily water consumption, total water requirement in growth season, etc.);

[0061] Obtain the number of each type of vegetation in each management area, calculate the distribution proportion of vegetation types, and obtain the growth stage of different vegetation types; according to the distribution proportion of vegetation types and the growth stage of different vegetation types, the baseline water requirement of different vegetation in different growth stages in the vegetation water requirement characteristic database is used to calculate the basic water requirement data of each management area; Taking a certain management area as an example, the system detects that there are two plants of Populus euphratica and Tamarix in the area, Populus euphratica accounts for 60%, and Tamarix accounts for 40%; Through the sensor, it is determined that Populus euphratica is in the growth period, and Tamarix is in the flowering period; The system retrieves the baseline water requirement of Populus euphratica in the growth period from the database, which is a liters / day / plant, and the baseline water requirement of Tamarix in the flowering period is b liters / day / plant; Combined with the respective distribution proportion and plant number, the basic water data of the management area is calculated as c liters / day.

[0062] Further, the growth stage of different vegetation types is obtained in the following way:

[0063] Image acquisition devices (such as high-definition cameras) are arranged in each management area, and different vegetation is photographed at a predetermined time interval (such as once a day) to obtain its RGB image; After gray scale normalization and noise filtering processing of the collected vegetation image, a directory storage structure is established according to the vegetation type, and a time series image data set strictly bound to the vegetation type is generated;

[0064] The pre-trained ResNet-50 convolutional neural network without full connection layer is used to extract the feature vector of each image in the time series data set, so as to obtain the appearance feature vector reflecting the appearance change of the vegetation. The appearance feature vector is arranged in time sequence to form a feature vector time series; The appearance feature vector includes leaf number feature vector (leaf node activation value), shape feature vector (contour curvature coefficient), reproductive feature vector (flower / fruit detection probability) and color distribution feature vector (HSV channel statistics);

[0065] For each type of vegetation, images of different growth stages such as long leaf period, dormancy period, flowering period, fruiting period, etc. are collected; the images are processed by using a ResNet-50 convolutional neural network with removed full connection layer to extract the baseline feature vectors of each growth stage; the mean value of the baseline feature vectors at the same growth stage is calculated to obtain the feature template vectors corresponding to different growth stages;

[0066] The similarity between the current time sequence feature vector and the feature template vector of the same vegetation at different growth stages is calculated (such as the cosine similarity between two feature vectors), and the growth stage corresponding to the highest similarity (such as long leaf period, dormancy period, flowering period, fruiting period, etc.) is determined as the current growth stage of the vegetation;

[0067] The physiological parameters of different vegetation in each management area are collected, and the collected physiological parameters are subjected to time series data analysis to identify the physiological state of the vegetation;

[0068] According to the physiological state of the vegetation, the basic water requirement data of each management area is adjusted to obtain the vegetation water requirement estimation data of each management area.

[0069] Further, the way of identifying the physiological state of the vegetation by time series data analysis of the collected physiological parameters includes:

[0070] The collected physiological parameter data is subjected to timestamp alignment, denoising, outlier removal and normalization processing to form standardized (timestamp and sampling frequency aligned, no abnormal value) vegetation physiological parameters; and the water stress index of the vegetation is calculated based on the standardized vegetation physiological parameters;

[0071] Based on the known rules of vegetation physiological processes (such as water transport, photosynthesis, nutrient metabolism, etc.), the vegetation physiological parameters are subjected to collaborative analysis to obtain the collaboration relationship (positive correlation or negative correlation) between the vegetation physiological parameters; for example, the collaboration relationship between leaf water content and vegetation water potential is positively correlated, the higher the leaf water content, the higher the vegetation water potential, but the collaboration relationship between leaf water content or vegetation water potential and transpiration rate is negatively correlated, the higher the transpiration rate, the lower the leaf water content or vegetation water potential;

[0072] Collaborative analysis is performed on leaf water content, stomatal conductance, transpiration rate, photosynthetic rate, chlorophyll content, and vegetation water potential to obtain the collaboration relationship between the vegetation physiological parameters; based on the collaboration relationship between the vegetation physiological parameters, a feature index is constructed, the constructed feature index is spliced to obtain a state feature vector; the state feature index includes hydraulic efficiency, photosynthetic water stress index, water consistency, stomatal-photosynthetic collaboration degree, and water use aggressiveness index; the above feature indexes are spliced to construct a state feature vector reflecting the physiological state of the vegetation;

[0073] The feature vector is dimensionally reduced by using a uniform manifold approximation and projection (UMAP) algorithm to obtain resonance frequency space coordinates (x, y, z) of the feature vector; and resonance intensity of the resonance frequency space coordinates is calculated to obtain resonance intensity of the feature vector;

[0074] The preset water stress index classification threshold is: healthy: [0, 0.3]; mild stress: (0.3, 0.5]; moderate stress: (0.5, 0.7]; severe stress: [0.7, 1.0]; the water stress index and the resonance intensity classification threshold are obtained through experiments; the preset resonance intensity classification threshold is: healthy: (0.8, 1.2]; mild stress: (0.6, 0.8]; moderate stress: (0.4, 0.6]; severe stress: [0, 0.4];

[0075] Based on the water stress index and the resonance intensity, and in combination with the preset water stress index and resonance intensity classification threshold, the physiological state of the vegetation is evaluated to obtain the physiological state of the vegetation; the physiological state includes healthy, mild stress, moderate stress and severe stress; for example, when the resonance intensity is 0.7 and the water stress index is 0.35, it indicates that the vegetation is in a mild stress state;

[0076] It should be noted that the water stress index calculation formula of the vegetation is: Wherein, PHS represents the water stress index; P n represents the measured photosynthetic rate; P n0 represents the photosynthetic rate of the vegetation in a healthy state; G s represents the measured stomatal conductance; G s0 represents the stomatal conductance of the vegetation in a healthy state; ψ represents the measured vegetation water potential; ψ0 represents the vegetation water potential of the vegetation in a healthy state; Chl represents the measured chlorophyll content; Chl0 represents the chlorophyll content of the vegetation in a healthy state;

[0077] The formula for calculating the hydraulic efficiency is: The formula for calculating the photosynthetic water stress index is: The formula for calculating the water use consistency is: 1-std(LWC, ψ, Tr); the formula for calculating the stomatal-photosynthetic coordination degree is: P n ×G s ; the formula for calculating the water use aggressivity index is: Wherein, LWC represents the leaf water content; Tr represents the transpiration rate; ε is a constant to avoid a denominator of 0, and the value is 0.000001; std() is a standard deviation function;

[0078] The formula for calculating the resonance intensity is: Wherein, (x, y, z) represents the resonance frequency space coordinates of the feature vector; (x0, y0, z0) represents the resonance frequency space coordinates of the ideal feature vector when the vegetation is in a healthy state, which is obtained based on the data of several target vegetation types in a healthy state through feature extraction, statistical modeling and dynamic calibration.

[0079] Further, the way of adjusting the basic water demand data of each management area according to the physiological state of the vegetation comprises:

[0080] Based on the physical meaning of the resonance intensity corresponding to the current physiological state of the vegetation and the water stress index (the resonance intensity reflects the synergy of various physiological parameters of the vegetation, and the water stress index reflects the water shortage degree of the vegetation); the response values of the resonance intensity and the water stress index are calculated in real time by using a segmented Sigmoid function, and are weighted and fused to obtain a water shortage factor; meanwhile, as a core factor affecting the physiological activities of the vegetation, the temperature directly affects the transpiration rate, stomatal conductance and other key parameters, and further affects the accuracy of the water shortage factor; for example, in a high-temperature environment (such as above 35℃), the transpiration rate of the vegetation is significantly higher than the average value under suitable temperature, which will cause the calculated water shortage factor to be too high (exceeding the actual demand); and in a low-temperature environment (such as below 10℃), the metabolic activity of the vegetation is weakened, and the transpiration rate is lower than the average value under suitable temperature, which will also cause the water shortage factor to be too low (lower than the actual demand); therefore, the temperature factor needs to be introduced to modify the water shortage factor to eliminate the deviation caused by the temperature difference; specifically, the water shortage factor is lowered by the temperature factor in high temperature (to avoid excessive water supply), and the water shortage factor is raised by the temperature factor in low temperature (to avoid insufficient water supply), and finally a water demand correction factor is generated, which can accurately reflect the actual water demand of the vegetation;

[0081] It should be noted that the function expression of the water demand correction factor is: τ = 1 + ((a1·f1(R) + a2·f2(PHS)) × γ(t)); wherein, τ represents the water demand correction factor; a1·f1(R) + a2·f2(PHS) represents the water shortage factor; γ(t) represents the temperature factor; R represents the resonance intensity, and PHS represents the water stress index; f1(R) represents the response function of the resonance intensity; f2(PHS) represents the response function of the water stress index; a1 and a2 represent weight coefficients, the sum of which is 1, and are dynamically adjusted with the physiological state of the vegetation; for example, when the vegetation is in a healthy state, the values are 0.7 and 0.3 respectively; when the vegetation is in a mild stress or moderate stress state, the values are 0.5 and 0.5 respectively; when the vegetation is in a severe stress state, the values are 0.3 and 0.7 respectively; the values are obtained by data fitting, and are different for different vegetation types; the function expression of the response function of the resonance intensity is: Wherein the values of r0 and r1 are 0.8 (lower limit of healthy interval) and 1.2 (upper limit of healthy interval) respectively; k1 and k2 represent the slope of stress state (control the growth rate of f1(R) when R decreases, the greater the value, the more sensitive) and the slope of over-coordination (control the decline rate of f1(R) when R increases) respectively; the function expression of the response function of the water stress index is: Wherein the values of PHS0 and PHS1 are 0.3 (upper limit of threshold of health) and 0.5 (upper limit of threshold of mild stress); k3 represents the slope parameter for controlling the steepness of the growth of f2(PHS) after PHS exceeds 0.3; the function expression of the temperature factor is: Wherein t represents the temperature when the vegetation physiological parameter is measured, t opt represents the optimum temperature for the physiological activity of vegetation; λ represents the temperature sensitivity coefficient, usually taking 20, for controlling the steepness of the curve, the smaller the value, the more obvious the change of the factor when the temperature deviates slightly; c represents the correction strength, taking the value range of 0.3-0.5, for controlling the maximum correction amplitude; sign() is the sign function, which is 1 when t-t opt is greater than 0, and -1 when it is less than 0.

[0082] Further, the method for correcting the vegetation water requirement estimation data of each management area by using the soil data comprises:

[0083] Obtaining the minimum soil water content threshold required for the growth of vegetation in each management area; if there are multiple types of vegetation in a management area, the soil water content required for the vegetation with the highest water content is taken as the minimum soil water content threshold required for the growth of vegetation in the management area;

[0084] Obtaining the salt tolerance threshold of different vegetation in the management area, selecting the vegetation with the worst salt tolerance (the lowest salt tolerance threshold), and setting 50%-70% of the salt tolerance threshold as the soil electrical conductivity threshold of the area (for example, if the salt tolerance threshold of cotton is 7.7 dS / m, the soil electrical conductivity threshold of the area can be set to 3.85-5.39 dS / m);

[0085] If the soil electrical conductivity is greater than the electrical conductivity corresponding to the salt tolerance threshold of the vegetation in the management area, the soil water resource data is recorded as 0;

[0086] If the soil electrical conductivity of the management area is less than the electrical conductivity corresponding to the salt tolerance threshold of the vegetation in the area, the soil water resource data directly provided by the soil in each management area is calculated based on the soil water content and the minimum soil water content required for the growth of vegetation;

[0087] If the calculated soil water resource data of each management area is greater than or equal to the estimated vegetation water demand data of each management area, it indicates that the soil water of the area itself can meet the vegetation growth, and no external water supplement is needed, and there is no water resource gap; otherwise, it indicates that external water supplement is needed, and the supplement amount is the difference between the two, which is the water resource gap data of the management area, and then the water resource gap data of each management area is obtained;

[0088] Based on the soil type, soil electrical conductivity and the above-mentioned pre-set soil electrical conductivity threshold value of the management area, the water quantity required for leaching soil salt to a safe range is calculated through a pre-set leaching water demand calculation model, and the leaching water quantity data of each management area is obtained.

[0089] The water resource gap data and the leaching water quantity data of each management area are added, and the comprehensive water resource demand data of each management area is obtained.

[0090] It should be noted that the expression of the pre-set leaching water demand calculation function is: Wherein, Q represents the calculated leaching water quantity; EC now represents the current soil electrical conductivity, which reflects the actual salt content of the current soil; EC aim represents the pre-set soil electrical conductivity threshold value; d represents the soil leaching depth (usually the depth of crop root layer 0.6-1.2m); θ fc represents the maximum water content of the soil after gravity drainage (reflecting the soil water storage capacity), which directly affects the total amount of salt migration with water; ρ represents the soil bulk density, unit: g / cm 3 ; represents the dry weight of unit volume of soil, which is used to convert the volume water content into mass-based salt content calculation and correct the salt storage difference of different compactness of soil (clay bulk density is greater than sand); η represents the leaching efficiency coefficient (dimensionless) which takes different coefficients according to the type of soil; EC out represents the electrical conductivity of leaching water, and the leaching water is fresh water.

[0091] Further, the generation mode of the water resource scheduling instruction includes:

[0092] The water demand and leaching water quality requirements of each management area (derived from the comprehensive water resource demand data) and the available water resource data of each water source point in the desert area are obtained; at the same time, the structure information and the maximum water transmission capacity of each pipe section of the water transmission pipe network are obtained; taking the minimization of water transmission cost as the optimization objective, setting the water supply capacity of the water source point, the transmission capacity of the pipe network and the irrigation water quality as the constraint conditions, a water resource scheduling optimization model is constructed; the water source point, the pipe network node and the management area are modeled as the nodes of the network flow diagram, and the water quantity allocation, the path selection and the time sequence scheduling are taken as the decision variables; when the water quality of the water source does not meet the irrigation requirements, water resource desalination treatment is carried out;

[0093] The water delivery cost calculation formula is: total cost = desalination cost + pump station energy consumption cost + pipe network operation cost + maintenance cost; wherein, the pump station energy consumption cost = water delivery flow rate x distance x time x electricity price; the pipe network operation cost is calculated according to the pipe segment length, flow rate and energy consumption coefficient per unit length; the maintenance cost is determined according to the total length of the pipe network and the maintenance cost per unit length; the constraint conditions specifically include: the total water output of each water source point does not exceed its supply capacity; the flow rate of each pipe network pipe segment does not exceed its maximum water delivery capacity; the actual water amount of each management area does not fall below its water demand; the actual water quality (such as the salt content after mixing) of each management area does not exceed its water quality threshold; the model construction is a common means for those skilled in the art, and the specific steps or processes of construction are not given here;

[0094] In the model solving stage, a linear programming or network flow optimization algorithm is used to calculate the optimal water supply amount of each water source point, the water delivery amount of each pipe network pipe segment, the actual water amount of each management area, and the specific water delivery path and delivery timing; and corresponding scheduling instructions are generated, which include the water supply amount of each water source point, the water delivery path distribution of the water delivery pipe network, and the water delivery amount and water delivery time of each management area water distribution point; the scheduling instructions are issued to the relevant valves, pump stations and pipe networks through an intelligent control system to realize automatic and accurate allocation;

[0095] In the scheduling execution process, the system continuously monitors the actual water supply amount of each node and the pipe network operation state; if it is found that the actual water supply amount or water quality deviates greatly from the model prediction, or that a pipe network failure, water source abnormality or other sudden situation occurs, the system will automatically re-collect data, dynamically adjust the optimization model and scheduling instructions, and realize closed-loop optimization and adaptive control; through this method, the water safety and water quality standard of each management area can be guaranteed, while the water delivery cost is minimized, and the water resource utilization efficiency and the intelligent level of scheduling are improved; for example, in a certain desert control area, the water demand of A, B and C management areas is 20, 30 and 25 cubic meters respectively, and the maximum allowable salt content is 1.0, 1.5 and 1.2 g / L respectively; the water supply amount of water source points S1 and S2 is 40 and 50 cubic meters respectively, and the salt content is 0.8 and 2.0 g / L respectively; the maximum water delivery capacity of the pipe network pipe segment is 30 cubic meters / hour; the optimization model automatically allocates S1 to supply A and B areas preferentially, S2 to supply C area, and mixes and delivers through the pipe network when necessary, to ensure that all areas meet the required water amount and the total water delivery cost is minimized; after the scheduling instructions are issued, the system continuously monitors the actual water amount and salt content of each area, and if it is found that the water quality of C area is too high, the allocation ratio and delivery path of S1 and S2 are automatically adjusted to ensure the dynamic optimization of scheduling.

[0096] The embodiment achieves the precision and intelligence breakthrough of desert water resource allocation by dividing the desert grid management area and synchronously acquiring heterogeneous parameters such as soil conductivity, vegetation water potential, stomatal conductance and temperature: it innovates the water demand prediction mechanism, dynamically classifies the evaluation based on the physiological state of the vegetation (healthy / mild / moderate / severe stress), generates a water demand correction factor based on real-time temperature data, dynamically quantifies the water demand response of high temperature transpiration surge and salt drought combined stress through time series analysis of the physiological parameter synchrony change, overcomes the response lag of traditional static model to the fluctuation of water use efficiency of vegetation to reduce irrigation redundancy; innovates the salinization control path, creates a leaching water demand calculation model of soil conductivity exceeding the salt tolerance threshold of vegetation, couples the soil type water conductivity characteristics and salt migration dynamics law, accurately calculates the leaching water quantity when the conductivity sensor triggers the critical early warning and includes the demand, synchronously achieves salt removal and invalid water consumption reduction to reverse the vegetation salt stress mortality; realizes the performance transition of global scheduling, builds a three-dimensional optimization model of water supply limit constraint, leaching demand constraint and water pipe network transmission capacity constraint, generates scheduling instructions with the minimum water transmission cost as the target, breaks through the limitations of single-point irrigation and water source island to compress the energy consumption and time delay of pipe network water transmission; effectively improves the problem of inefficient allocation of desert water resources.

[0097] Embodiment 2

[0098] The part not described in detail in the embodiment can be referred to the description of the embodiment 1. Figure 2 As shown in the figure, the desert area water resource intelligent allocation system based on Internet of Things provided by the embodiment comprises:

[0099] The data acquisition module is configured to divide the desert area into a plurality of management areas, and acquire soil data, vegetation species, vegetation physiological parameters in each management area, and available water resource data of each water source point in the desert area.

[0100] The water demand calculation module is configured to determine the vegetation water demand estimation data of each management area based on the vegetation species and the vegetation physiological parameters in each management area.

[0101] The water demand adjustment module is configured to correct the vegetation water demand estimation data of each management area by using the soil data, and generate comprehensive water resource demand data of each management area.

[0102] The scheme generation module is configured to generate a water resource scheduling scheme of the desert area based on the comprehensive water resource demand data of each management area and the available water resource data of each water source point.

[0103] The above merely describes preferred embodiments of the present application, and is not intended to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, the technical solutions recorded in the foregoing embodiments can be modified or some technical features can be replaced equivalently by those skilled in the art, without departing from the spirit and principle of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall fall within the scope of the present application.

[0104] It should be noted that, in this document, the terms "comprises", "comprising", or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without more limitations, an element preceded by "comprises... a" does not exclude the existence of additional identical elements in the process, method, article, or apparatus that comprises the stated element.

[0105] In the description of the specification, the description of the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in one or more embodiments or examples.

[0106] The formulas in the specification are dimensionless numerical calculations, the formulas are obtained by collecting a large amount of data to simulate the most recent real situation, and the preset parameters and threshold values in the formulas are set by those skilled in the art according to the actual situation.

[0107] Although the embodiments of the present application have been shown and described, those skilled in the art can understand that various changes, modifications, replacements and variations can be made to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the claims and their equivalents.

Claims

1. A method for intelligent allocation of water resources in desert areas based on the Internet of Things, characterized in that, include: The desert area was divided into several management areas, and soil data, vegetation types and vegetation physiological parameters were obtained in each management area, as well as available water resource data of each water source in the desert area. Based on the vegetation species and physiological parameters within each management area, the estimated vegetation water demand data for each management area were determined. The methods for obtaining this estimated vegetation water demand data for each management area included: The distribution ratio of different vegetation species in each management area was statistically analyzed, and the growth stage of different vegetation species was obtained. Based on the distribution ratio of vegetation types and the growth stage of different vegetation types, the basic water demand data is matched from the preset vegetation water demand characteristic database to determine the basic water demand data of each management area. Collect vegetation physiological parameters of different vegetation in each management area, and perform time-series data analysis on the collected vegetation physiological parameters to identify the physiological state of the vegetation. Based on the physiological state of the vegetation, the basic water demand data of each management area are adjusted to obtain the estimated water demand data of the vegetation in each management area. Using the soil data, the estimated vegetation water demand data for each management area are corrected to generate comprehensive water resource demand data for each management area. The methods for correcting the estimated vegetation water demand data for each management area include: The minimum soil moisture content and salt tolerance threshold required for vegetation growth in each management area are preset. If the soil electrical conductivity of the management area is greater than or equal to the electrical conductivity corresponding to the salt tolerance threshold of the vegetation in the corresponding management area, then the available soil water resources data for that management area are zero. If the soil electrical conductivity of the management area is less than the electrical conductivity corresponding to the salt tolerance threshold of the vegetation in the corresponding management area, then the available soil water resources data for the management area are calculated based on the difference between the soil moisture content and the preset minimum soil moisture content. By comparing the soil water resources data of each management area with the estimated vegetation water demand data of each management area, the water resource gap data of each management area can be obtained. Based on the salt tolerance threshold of vegetation in each management area, soil electrical conductivity thresholds are set for each management area; based on soil type, soil electrical conductivity, and the preset soil electrical conductivity thresholds, the leaching water volume data for each management area is obtained through a preset leaching water demand calculation function. The water resource gap data and the rinsing water volume data of each management area are added together to obtain the comprehensive water resource demand data of each management area; Based on the comprehensive water resource demand data of each management area and the available water resource data of each water source, a water resource allocation plan for desert areas is generated.

2. The method for intelligent allocation of water resources in desert areas based on the Internet of Things as described in claim 1, characterized in that, Soil data include soil type, soil electrical conductivity, and soil moisture content; vegetation physiological parameters include leaf water content, stomatal conductance, transpiration rate, photosynthetic rate, chlorophyll content, vegetation water potential, and ambient temperature.

3. The method for intelligent allocation of water resources in desert areas based on the Internet of Things as described in claim 1, characterized in that, The methods for obtaining the growth stages of different vegetation species include: Images of each vegetation species are collected at preset time intervals. The collected vegetation images are preprocessed to form a time-series image dataset bound to each vegetation species. Feature extraction is performed on the images in the time-series image dataset to generate a time-series sequence of appearance feature vectors for each vegetation species; The temporal sequence of the appearance feature vector is input into a ResNet-50 convolutional neural network with the fully connected layers removed to obtain a temporal feature vector characterizing the vegetation growth dynamics. The feature template vectors of different growth stages of vegetation are obtained, and the temporal feature vectors are matched with the feature template vectors of different growth stages of vegetation to determine the current growth stage of the vegetation species.

4. The method for intelligent allocation of water resources in desert areas based on the Internet of Things according to claim 3, characterized in that, The methods for identifying the physiological state of vegetation by performing time-series data analysis on the collected vegetation physiological parameters include: The collected physiological parameter data were processed by timestamp alignment, noise reduction, outlier removal and normalization to form standardized vegetation physiological parameters, and the vegetation water stress index was calculated based on the standardized vegetation physiological parameters. Synergistic analysis was conducted on leaf water content, stomatal conductance, transpiration rate, photosynthetic rate, chlorophyll content, and vegetation water potential to obtain the cooperative relationship among vegetation physiological parameters. Based on the cooperative relationship among vegetation physiological parameters, feature indicators were constructed, and the constructed feature indicators were spliced ​​to obtain a state feature vector. The dimensionality of the state feature vector is reduced by using the uniform manifold approximation and projection algorithm to obtain the resonant frequency spatial coordinates of the state feature vector; the resonant intensity is calculated from the resonant frequency spatial coordinates to obtain the resonant intensity of the state feature vector. Based on the water stress index and resonance intensity, and combined with the preset grading thresholds of the water stress index and resonance intensity, the physiological state of vegetation is assessed, and the physiological state of vegetation is obtained.

5. The method for intelligent allocation of water resources in desert areas based on the Internet of Things according to claim 4, characterized in that, The methods for adjusting the basic water requirement data for each management area based on the physiological state of the vegetation include: Based on the physiological state of vegetation and its corresponding resonance intensity and the vegetation water stress index, a water demand correction factor is generated in combination with the ambient temperature. The basic water demand data of vegetation is updated using the water demand correction factor, thereby obtaining the vegetation water demand prediction data for each management area.

6. The method for intelligent allocation of water resources in desert areas based on the Internet of Things according to claim 1, characterized in that, The methods for generating water resource allocation plans include: A water resource scheduling optimization model is constructed. The optimization objective of the water resource scheduling optimization model is to minimize the water transmission cost. The constraints are the available water supply capacity of each water source and the maximum transmission capacity of each pipeline in the water transmission network. Based on the available water resource data of each water source, the comprehensive water resource demand data of each management area, the optimization objectives and constraints, the water resource scheduling optimization model is solved using a network flow optimization algorithm to generate a water resource scheduling scheme.

7. An Internet of Things (IoT)-based intelligent water resource allocation system for desert areas, used to implement the IoT-based intelligent water resource allocation method for desert areas as described in any one of claims 1 to 6, characterized in that, include: The data acquisition module is used to divide the desert area into several management areas and acquire soil data, vegetation types, vegetation physiological parameters, and available water resource data of each water source in the desert area. The water demand calculation module is used to determine the estimated water demand data of vegetation in each management area based on the vegetation species and vegetation physiological parameters in each management area. The water demand adjustment module is used to correct the vegetation water demand forecast data of each management area using the soil data, and generate comprehensive water resource demand data for each management area. The scheme generation module is used to generate water resource scheduling schemes for desert areas based on comprehensive water resource demand data of each management area and available water resource data of each water source.

Images