CN104521699A

CN104521699A - Field intelligent irrigation on-line control management method

Info

Publication number: CN104521699A
Application number: CN201410655632.1A
Authority: CN
Inventors: 马建琴; 张振伟; 刘雪梅; 彭高辉; 郝秀平; 刘蕾; 丁泽霖; 杨雪颖; 张富; 王婧; 王伟; 杨宁
Original assignee: North China University of Water Resources and Electric Power
Current assignee: North China University of Water Resources and Electric Power
Priority date: 2014-11-18
Filing date: 2014-11-18
Publication date: 2015-04-22

Abstract

The invention discloses an online control and management method for intelligent irrigation in the field. A small automatic weather station collects meteorological data, and soil water environment monitoring equipment measures the soil moisture content in the root zone of crops in real time, and imports and stores it into a database; the client logs in to the server online. Real-time intelligent irrigation system for farmland; management of field soil information, crop information management, real-time monitoring of farmland, real-time field irrigation simulation and irrigation forecast, field information query, and information statistics. The invention can perform remote online irrigation real-time guidance. Based on the network, the data collected in real time by the data acquisition system is analyzed, summarized, processed and other secondary processing, the real-time transpiration and evaporation of the crops are calculated, and the real-time insufficient irrigation simulation, irrigation forecast and irrigation management of the crops are performed online. Through model simulation, parameters The internal processing of the system such as simulation can accurately determine whether crops need to be irrigated, the amount of irrigation water, and the time of irrigation.

Description

Field intelligent irrigation online control management method

技术领域 technical field

本发明涉及农田智能灌溉管理技术领域，特别是涉及一种田间智能灌溉在线控制管理方法。 The invention relates to the technical field of farmland intelligent irrigation management, in particular to an online control and management method for field intelligent irrigation.

背景技术 Background technique

当前我国水资源严重短缺，水资源供需矛盾日益突出，农业用水尤其紧张。因此，加强我国农田灌溉管理，实现农田灌溉的精准化、自动化、智能化和现代化，切实提高农业水资源的高效、节水利用水平，不断完善农业水资源的优化配置制度，全面提高水资源利用效率、实现水资源的可持续发展已成为当务之急；也是以农田灌溉的现代化支撑并促进水利现代化建设、实现农业现代化的迫切要求。 At present, there is a serious shortage of water resources in my country, and the contradiction between supply and demand of water resources is becoming more and more prominent, especially for agricultural water use. Therefore, it is necessary to strengthen my country's farmland irrigation management, realize the precision, automation, intelligence and modernization of farmland irrigation, effectively improve the efficient and water-saving utilization of agricultural water resources, continuously improve the optimal allocation system of agricultural water resources, and comprehensively improve the utilization of water resources. Efficiency and sustainable development of water resources have become a top priority; it is also an urgent requirement to support and promote the modernization of water conservancy and realize the modernization of agriculture with the modernization of farmland irrigation.

我国农业水资源紧缺对粮食安全造成了很大威胁，在作物生长期如何充分、有效地利用当地降水，缓解农业水资源紧缺，切实提高作物产量，是目前急需解决的难点问题之一。 The shortage of agricultural water resources in my country poses a great threat to food security. How to fully and effectively use local precipitation during the crop growth period to alleviate the shortage of agricultural water resources and effectively increase crop yields is one of the difficult problems that need to be solved urgently.

传统的农田灌溉用水管理一般是根据不同水文年的配水方案预先制定灌溉制度，然而在生产实践中，由于实际年型与预测的水文年型之间存在着差异，传统的用水管理很难适应瞬息万变的天气条件。在目前水资源紧缺、农业供水形势日益严峻、灌溉管理水平比较低的情况下，结合未来时段的天气预报进行农田灌溉用水动态管理具有重要意义。随着计算机网络技术与监测技术的快速发展，将农业、气象等实时数据作为输入，探讨作物实时灌溉制度和预报技术，解决非充分灌溉条件下作物实时灌溉的科学问题，也正在成为一种新的研究内容。 Traditional irrigation water management for farmland is generally based on pre-established irrigation systems based on water allocation plans for different hydrological years. However, in production practice, due to the differences between the actual year pattern and the predicted hydrological year pattern, it is difficult for traditional water management to adapt to changing conditions. weather conditions. In the current situation of shortage of water resources, increasingly severe agricultural water supply situation, and relatively low level of irrigation management, it is of great significance to dynamically manage farmland irrigation water in combination with weather forecasts in the future. With the rapid development of computer network technology and monitoring technology, it is becoming a new method to use real-time data such as agriculture and meteorology as input to explore real-time crop irrigation systems and forecasting technologies, and to solve scientific problems of real-time crop irrigation under insufficient irrigation conditions. research content.

发明内容 Contents of the invention

本发明针对我国缺乏可用于实际操作的远程在线控制技术和农业高效用水智能决策技术问题，提供一种田间智能灌溉在线控制管理方法。 The invention provides an online control and management method for field intelligent irrigation in view of the lack of remote online control technology and intelligent decision-making technology for efficient agricultural water use in my country that can be used in actual operations.

技术方案：一种田间智能灌溉在线控制方法，包括以下步骤： Technical solution: an online control method for field intelligent irrigation, comprising the following steps:

步骤一，小型自动气象站采集气象数据，土壤水环境监测设备实时测定作物根区的土壤含水率，采集和测定的数据导入并存储到位于互联网上的服务器的数据库中； Step 1, the small automatic weather station collects meteorological data, the soil water environment monitoring equipment measures the soil moisture content in the root zone of the crop in real time, and the collected and measured data are imported and stored in the database of the server located on the Internet;

步骤二，客户端在线登录位于所述服务器上的农田实时智能灌溉软件系统； Step 2, the client logs in to the farmland real-time intelligent irrigation software system located on the server;

步骤三，所述农田实时智能灌溉软件系统调用所述数据库中的数据，根据客户端操作进行田间土壤信息的管理、农作物信息管理、农田实时监控、田间实时灌溉模拟与灌溉预报、田间信息查询、信息统计。 Step 3, the farmland real-time intelligent irrigation software system invokes the data in the database, and performs field soil information management, crop information management, farmland real-time monitoring, field real-time irrigation simulation and irrigation forecast, field information query, information statistics.

所述田间实时灌溉模拟与灌溉预报，在短期气象预报的基础上，根据灌区各田块实际土壤水分、作物实际蒸腾蒸发量、作物生长状况，通过建立土壤墒情预测模型、作物需水预测模型、实时灌溉预报模型、作物系数实时修正模型，实时预测作物在预报期是否需要灌水以及灌水的时间、灌水量，并在每个计算时段结束时，用土壤含水率实测值对土壤含水率的模拟值、作物系数进行修正。 The real-time irrigation simulation and irrigation forecast in the field is based on the short-term weather forecast, according to the actual soil moisture of each field in the irrigated area, the actual transpiration and evaporation of the crops, and the growth status of the crops, by establishing a soil moisture prediction model, a crop water demand prediction model, Real-time irrigation forecasting model, crop coefficient real-time correction model, real-time prediction of whether crops need to be irrigated during the forecast period, the time of irrigation, and the amount of irrigation, and at the end of each calculation period, the simulated value of soil moisture content is compared with the measured value of soil moisture content , The crop coefficient is corrected.

一种田间智能灌溉在线控制管理方法，包括用于监测土壤湿润层含水量W的土壤水分检测单元，用于监测实际降雨量P的降雨检测单元，以及灌溉系统，还包括土壤墒情预测模型、作物实时需水量的模糊聚类预测模型和在线实时灌溉预报及控制模型； A field intelligent irrigation online control management method, including a soil moisture detection unit for monitoring the moisture content W of the soil wet layer, a rainfall detection unit for monitoring the actual rainfall P, and an irrigation system, and also includes a soil moisture prediction model, Fuzzy clustering forecasting model of crop real-time water demand and online real-time irrigation forecasting and control model;

所述土壤墒情的实时预测模型：W_i＝W_i-1+P_0i+W_Ti-ET_i+M_i+K_i The real-time prediction model of the soil moisture content: W _i =W _i-1 +P _0i +W _Ti -ET _i +M _i +K _i

其中：W_i-1——第i天初始计划湿润层的土壤含水量； Among them: W _i-1 ——the soil water content of the initial planned wet layer on the i-th day;

W_i——第i天结束时计划湿润层的土壤含水量； W _i —soil moisture content of the planned wetting layer at the end of the i-th day;

P_0i——第i天的有效降雨量； P _0i - the effective rainfall on the i-th day;

W_Ti——第i天由计划湿润层增加而增加的水量； W _Ti ——the amount of water increased by the increase of the planned wetting layer on the i-th day;

ET_i——第i天的作物需水量； ET _i ——crop water requirement on day i;

M_i——第i天的灌水量； M _i ——Irrigation amount on the i-th day;

K_i——第i天的地下水补给量； K _i — groundwater recharge on the i-th day;

所述作物实时需水量的模糊聚类预测模型：ET＝ET₀'·K_c·K_w＝s·ET₀·K_c·K_w； The fuzzy clustering prediction model of the real-time water demand of the crops: ET=ET ₀ '·K _c ·K _w =s·ET ₀ ·K _c ·K _w ;

其中：ET_i—第i天的作物需水量； Where: ET _i — crop water requirement on day i;

ET_0i—第i天的参考作物需水量； ET _0i — reference crop water requirement on day i;

K_ci—第i天的作物系数； K _ci — crop coefficient on the i-th day;

K_wi—实行非充分灌溉时第i天的土壤水分修正系数； K _wi —Soil moisture correction coefficient on the i-th day when non-full irrigation is implemented;

s-不同天气类型的聚类中心， s - cluster centers of different weather types,

得到以天为时段的作物在线实时灌溉预报模型： Get the crop online real-time irrigation forecast model with days as the time period:

M_i＝W_i-1+P_0i+W_Ti-ET_i+K_i-W_i M _i ＝W _i-1 +P _0i +W _Ti -ET _i +K _i -W _i

所述在线实时灌溉预报及控制模型根据作物实时灌水量M_i控制灌溉系统对作物实时进行灌溉。 The online real-time irrigation forecast and control model controls the irrigation system to irrigate the crops in real time according to the real-time irrigation water volume M _i of the crops.

还包括作物计划湿润层含水率的逐日递推预测模型，若研究区域的地下水埋深较深，地下水对作物补给忽略不计时，则以天为预报时段的作物计划湿润层的水量平衡模型为： It also includes a day-by-day recursive prediction model for the moisture content of the crop planning wetting layer. If the buried depth of the groundwater in the study area is relatively deep, and the groundwater is negligible for crop replenishment, the water balance model of the crop planning wetting layer with days as the forecast period is:

W_i＝W_i-1+P_0i+W_Ti-ET_i+M_i； W _i =W _i-1 +P _0i +W _Ti -ET _i +M _i ;

其中，第i天初始、结束时计划湿润层土壤含水量如下： Among them, the soil water content of the planned wetting layer at the beginning and end of the i-th day is as follows:

W_i-1＝1000nH_i-1θ_i-1 W _i-1 ＝1000nH _i-1 θ _i-1

W_i＝1000nH_iθ_i W _i ＝1000nH _i θ _i

其中：H_i-1—第i天初始计划湿润层深； Among them: H _i-1 — initial planned wet layer depth on the i-th day;

H_i—第i天结束时的计划湿润层深； H _i —planned wetted layer depth at the end of day i;

θ_i-1—第i天初始土壤含水率，以占土壤体积百分比计； θ _i-1 —Initial soil moisture content on the i-th day, expressed as a percentage of soil volume;

θ_i—第i天结束时的土壤含水率，以占土壤体积百分比计； θ _i —soil moisture content at the end of the i-th day, expressed as a percentage of soil volume;

n—土壤孔隙率； n—soil porosity;

第i时段内因计划湿润层的增加而增加的水量W_Ti； The amount of water W _Ti increased due to the increase of the planned wetting layer in the i-th period;

W_Ti＝1000(H_i-H_i-1)·n·θ_deep W _Ti ＝1000(H _i -H _i-1 )·n·θ _deep

其中：θ_deep—深层土壤含水率； Where: θ _deep — deep soil moisture content;

可得出作物计划湿润层含水率的逐日递推预测模型： The daily recursive prediction model of the moisture content of the crop plan wetting layer can be obtained:

${θ θ}_{i i} = = \frac{{H h}_{i i - - 11}}{{H h}_{i i}} {θ θ}_{i i - - 11} - - (({ET ET}_{i i - - 11} - - {P P}_{00 i i - - 11} - - {W W}_{Ti Ti} - - {M m}_{i i - - 11})) / / ((10001000 n no {H h}_{i i})) . .$

按照作物计划湿润层含水率的逐日递推预测模型，从作物的种植日期开始，逐日对作物根区的土壤墒情进行预测，直至作物收获，从而进行作物整个生育期内土壤水分状况的逐日递推预测。 According to the day-by-day recursive prediction model of the moisture content of the wet layer in the crop plan, the soil moisture in the root zone of the crop is predicted day by day from the planting date of the crop until the crop is harvested, so as to carry out the day-by-day recursion of the soil moisture status during the entire growth period of the crop predict.

还包括作物需水量ET₀的模糊聚类预测方法：对不同天气类型下历史长系列ET₀进行模糊聚类，得到不同天气类型的聚类中心s，进行ET₀的自修正：ET′₀＝s·ET₀，达到更加精确的预测未来天气情景下作物需水量； Also includes the fuzzy clustering prediction method of crop water demand ET ₀ : carry out fuzzy clustering on the historical long series ET ₀ under different weather types, obtain the cluster centers s of different weather types, and perform self-correction of ET ₀ : ET′ ₀ = s·ET ₀ , to achieve a more accurate prediction of crop water demand under future weather scenarios;

所述聚类中心s的方法确定： The method of clustering center s is determined:

设x_ij为不同天气类型下的ET₀值，其中，i表示天气类型，i＝1，2，3，4，分别表示晴、云、阴、雨；j表示旬，j＝1，2，……，36；由ET₀分类统计表可得全年各旬不同天气类型下多年平均ET₀的特征值矩阵： Let x _ij be the ET ₀ value under different weather types, wherein, i represents the weather type, i=1, 2, 3, 4, respectively represent sunny, cloud, overcast, rain; j represents ten days, j=1, 2, ..., 36; from the ET ₀ classification statistics table, the eigenvalue matrix of the multi-year average ET ₀ under different weather types in each ten-day of the year can be obtained:

${x x}_{ij ij} = = [\begin{matrix} {x x}_{1111} & {x x}_{1212} & . . . . . . & {x x}_{11 j j} \\ {x x}_{21 twenty one} & {x x}_{21 twenty one} & . . . . . . & {x x}_{22 j j} \\ . . . . . . & . . . . . . & . . . . . . & . . . . . . \\ {x x}_{i i 11} & {x x}_{i i 22} & . . . . . . & {x x}_{ij ij} \end{matrix}]$

各旬ET₀特征值的相对隶属度矩阵： The relative membership degree matrix of ET ₀ eigenvalues in each ten-day period:

其中：maxx_ij——j旬ET₀特征值的最大值； Among them: maxx _ij - the maximum value of the eigenvalue of ET ₀ in j days;

则j旬第i种天气类型下的相对隶属度矩阵为： Then the relative membership degree matrix under the i-th weather type in ten days is:

$R R = = {r r}_{ij ij} = = [\begin{matrix} {r r}_{1111} & {r r}_{1212} & . . . . . . & {r r}_{11 j j} \\ {r r}_{21 twenty one} & {r r}_{22 twenty two} & . . . . . . & {r r}_{22 j j} \\ . . . . . . & . . . . . . & . . . . . . & . . . . . . \\ {r r}_{i i 11} & {r r}_{i i 22} & . . . . . . & {r r}_{ij ij} \end{matrix}]$

设i种天气类型下ET₀的模糊聚类中心矩阵为： Let the fuzzy cluster center matrix of ET ₀ under i weather types be:

$S S = = (({S S}_{ih i h})) = = [\begin{matrix} {S S}_{1111} & {S S}_{1212} \\ {S S}_{21 twenty one} & {S S}_{22 twenty two} \\ . . . . . . & . . . . . . \\ {S S}_{i i 11} & {S S}_{i i 22} \end{matrix}]$

其中：s_ih——为分类模式h下i种天气类型下ET₀的特征值规格化数，0≤s_ih≤1； Among them: s _ih —— is the normalized number of eigenvalues of ET ₀ under i weather types under the classification mode h, 0≤s _ih ≤1;

h＝1，2，…，c； h=1,2,...,c;

假设按对晴天ET₀的相对隶属度进行分类，即c＝2，则其模糊聚类矩阵为： Assuming that the classification is carried out according to the relative membership degree of the sunny day ET ₀ , that is, c=2, then its fuzzy clustering matrix is:

$U u = = (({u u}_{hj hj})) = = [\begin{matrix} {u u}_{1,1 1,1} & {u u}_{1,2 1,2} & . . . . . . & {u u}_{1,36 1,36} \\ {u u}_{2,1 2,1} & {u u}_{2,1 2,1} & . . . . . . & {u u}_{2,36 2,36} \end{matrix}]$

且满足条件： And meet the conditions:

$\begin{matrix} 00 \leq \leq {u u}_{hj hj} \leq \leq 11 \\ {Σ Σ}_{h h = = 11}^{c c} {u u}_{hj hj} = = 00 \\ {Σ Σ}_{j j = = 11}^{n no} {u u}_{hj hj} > > 00 \end{matrix}\}$

则j旬样本与不同天气类型ET₀聚类中心h间的差异，可用广义欧氏权距离表示，即 Then the difference between the j-day sample and the ET ₀ cluster center h of different weather types can be expressed by the generalized Euclidean weighted distance, namely

$| | | | {w w}_{i i} (({r r}_{j j} - - {s the s}_{h h})) | | | | = = {{{Σ Σ}_{i i = = 11}^{m m} [[{w w}_{i i} {(({r r}_{ij ij} - - {s the s}_{ih i h}))}^{22}]]}}^{\frac{11}{22}}$

其中：i——天气类型总数； Where: i—the total number of weather types;

以u_hj为权重得到样本j与类别h间的加权广义欧氏权距离： The weighted generalized Euclidean weighted distance between sample j and category h is obtained by taking u _hj as weight:

F(u_hj)＝u_hj||w_i(r_j-s_h)||， F(u _hj )＝u _hj ||w _i (r _j -s _h )||,

计算出聚类中心s，得到ET₀的修正值ET₀'。 Calculate the cluster center s, and get the corrected value ET ₀ ' of ET ₀ .

最优模糊聚类中心矩阵s求解步骤具体如下： The steps to solve the optimal fuzzy clustering center matrix s are as follows:

(1)给定u_hj与s_ih所要求满足的计算精度ε₁、ε₂； (1) Given the calculation accuracy ε ₁ and ε ₂ required by u _hj and _sih ;

(2)假设一个满足约束条件u_hj且元素不全部相等的初始模糊聚类矩阵 (2) Assume an initial fuzzy clustering matrix that satisfies the constraints u _hj and the elements are not all equal

(3)代入对应的初始模糊聚类中心矩阵 (3) substitute The corresponding initial fuzzy clustering center matrix

${S S}_{ih i h}^{00} = = \frac{{Σ Σ}_{j j = = 11}^{n no} {u u}_{hj hj}^{00^{22}} {w w}_{i i}^{22} {r r}_{ij ij}}{{Σ Σ}_{j j = = 11}^{n no} {u u}_{hj hj}^{00^{22}} {w w}_{i i}^{22}}$

(4)代入求一次近似模糊聚类矩阵 (4) substitute Find an approximate fuzzy clustering matrix

${u u}_{hj hj}^{11} = = \frac{11}{{Σ Σ}_{h h = = 11}^{c c} \frac{{Σ Σ}_{i i = = 11}^{m m} {[[{w w}_{i i} - - (({r r}_{ij ij} - - {s the s}_{ih i h}^{00}))]]}^{22}}{{Σ Σ}_{i i = = 11}^{m m} {[[{w w}_{i i} - - (({r r}_{ij ij} - - {s the s}_{ih i h}))]]}^{22}}}$

(5)代入求一次近似模糊聚类中心矩阵 (5) substitute Finding an approximate fuzzy clustering center matrix

(6)逐个比较矩阵和以及矩阵的对应元素，若且则迭代结束，为最优聚类中心矩阵s，否则重复(3)至(5)的步骤直至满足精度要求，已有研究已证明了迭代的收敛性；经模糊聚类计算，可以得到云、阴、雨天气情况下ET₀对于晴天ET₀的模糊聚类中心。 (6) Compare the matrices one by one and and the matrix The corresponding elements of , if and then the iteration ends, is the optimal cluster center matrix s, otherwise Repeat the steps from (3) to (5) until the accuracy _requirements are met. Existing studies have proved the convergence of iterations; through fuzzy clustering calculations, the fuzziness of ET 0 in cloudy, overcast and rainy weather to ET ₀ in sunny days can be obtained cluster center.

进一步，还对预报精度的分析方法： Further, the analysis method of forecast accuracy is also:

采用绝对累计偏差(ABE)、均方根误差(RMSE)、相对误差(RE)、决定系数(R²)和认同系数(IA)等指标来评定预报精度； The absolute cumulative deviation (ABE), root mean square error (RMSE), relative error (RE), coefficient of determination (R ² ) and coefficient of agreement (IA) are used to evaluate the forecast accuracy;

$ABE ABE = = {Σ Σ}_{k k = = 11}^{n no} | | {x x}_{k k} - - {y the y}_{k k} | |$

$RMSE RMSE = = \sqrt{\frac{{Σ Σ}_{k k = = 11}^{m m} {(({x x}_{k k} - - {y the y}_{k k}))}^{22}}{n no}}$

$RE RE = = \frac{RMSE RMSE}{\overset{&OverBar; &OverBar;}{x x}}$

${R R}^{22} = = \frac{{[[Σ Σ (({x x}_{k k} - - \overset{&OverBar; &OverBar;}{x x})) (({y the y}_{k k} - - \overset{&OverBar; &OverBar;}{y the y}))]]}^{22}}{Σ Σ {(({x x}_{k k} - - \overset{&OverBar; &OverBar;}{x x}))}^{22} Σ Σ {(({y the y}_{k k} - - \overset{&OverBar; &OverBar;}{y the y}))}^{22}}$

$IA IA = = 11 - - \frac{{Σ Σ}_{k k = = 11}^{n no} {(({x x}_{k k} - - {y the y}_{k k}))}^{22}}{{Σ Σ}_{k k = = 11}^{n no} {((| | {x x}_{k k}^{' '} | | + + | | {y the y}_{k k}^{' '} | |))}^{22}}$

${x x}_{k k}^{' '} = = {x x}_{k k} - - \overset{&OverBar; &OverBar;}{x x}$

${y the y}_{k k}^{' '} = = {y the y}_{k k} - - \overset{&OverBar; &OverBar;}{y the y}$

以上各式中x_k为预测值，y_k为实际值，k＝1，2，3，4，5……，n；分别为预测值序列和实际值序列的平均值，n为预测值和实际值序列的样本数。 Among the above formulas, x _k is the predicted value, y _k is the actual value, k=1, 2, 3, 4, 5..., n; are the mean values of the predicted value sequence and the actual value sequence, respectively, and n is the sample number of the predicted value sequence and the actual value sequence.

预测结果精度评价标准 Accuracy Evaluation Criteria for Prediction Results

进一步，根据作物来水条件进行充分或非充分灌溉，当来水量很大，足以满足灌溉需求时，灌水量M_i为： Further, sufficient or insufficient irrigation is carried out according to the incoming water conditions of the crops. When the incoming water is large enough to meet the irrigation demand, the amount of irrigation water M _i is:

M_i＝1000·n·H_i(1-θ_i)·θ_max M _i ＝1000·n·H _i (1-θ _i )·θ _max

其中：θ_i—第i天(阶段)的初始土壤含水率； Among them: θ _i — the initial soil moisture content of the i-th day (stage);

H_i—第i天(阶段)的作物计划湿润层深度； H _i —the depth of the crop plan wet layer on the i-th day (stage);

当来水不足或者水资源量紧缺时，对作物进行非充分灌溉；灌水量M_i为： When the inflow of water is insufficient or the amount of water resources is in short supply, the crops are under-irrigated; the amount of irrigation water M _i is:

M_i＝1000·n·H_i(θ_c1-θ_i)·θ_max M _i ＝1000·n·H _i (θ _c1 -θ _i )·θ _max

其中：θ_c1—灌溉后所要达到的土壤含水率，非充分灌溉根据作物耐受度确定具体取值进一步，土壤水分实时修正系数K_wi的修改方法为： Among them: θ _c1 — the soil moisture content to be achieved after irrigation, and the specific value of insufficient irrigation is determined according to the crop tolerance. Further, the modification method of the real-time correction coefficient K _wi of soil moisture is:

${K K}_{wi wi} = = \{\begin{matrix} ln ln ((11 + + 100100 \frac{{θ θ}_{i i}}{{θ θ}_{max max}})) / / ln ln 101101 & {θ θ}_{c c 22} \leq \leq {θ θ}_{i i} < < {θ θ}_{c c 11} \\ α α \cdot &Center Dot; exp exp [[(({θ θ}_{i i} - - {θ θ}_{c c 22}))]] / / {θ θ}_{c c 22} & {θ θ}_{i i} < < {θ θ}_{c c 22} \end{matrix}$

其中：θ_i—第i天土壤含水率，以占土壤体积百分比计； Among them: θ _i —soil moisture content on the i-th day, in percentage of soil volume;

θ_max—田间持水率，以占土壤体积百分比计； θ _max —Field water holding capacity, expressed as a percentage of soil volume;

θ_c1—非充分灌溉适宜土壤水分上限指标，以占田间持水率θ_max的百分数表示，研究中分别按不同的实验方案进行确定； θ _c1 — the upper limit index of soil moisture suitable for insufficient irrigation, expressed as a percentage of field water holding capacity θ _max , determined according to different experimental schemes in the research;

θ_c2—非充分灌溉适宜土壤水分下限指标，以占θ_max的百分数表示；研究中分别按不同的实验方案进行确定； θ _c2 —The lower limit index of soil moisture suitable for insufficient irrigation, expressed as a percentage of θ _max ; it was determined according to different experimental schemes in the research;

α—经验系数，同一作物为固定值。 α—experience coefficient, fixed value for the same crop.

进一步，所述作物系数K_ci初值的确定，在首次运行系统时，确定作物系数K_ci的初值，K_ci的计算初值采用随作物生育期累计天数i逐日变化的计算方法： Further, the determination of the initial value of the crop coefficient K _ci is to determine the initial value of the crop coefficient K _ci when the system is run for the first time, and the calculation of the initial value of K _ci adopts a calculation method that changes day by day with the cumulative days i of the crop growth period:

${K K}_{ci ci} = = \{\begin{matrix} 7.346 7.346 {((i i / / I I))}^{22} - - 1.606 1.606 ((i i / / I I)) + + 0.0972 0.0972 & i i = = I I \leq \leq 0.058 0.058 \\ - - 3.463 3.463 ln ln ((i i / / I I)) - - 0.1909 0.1909 & i i / / I I &GreaterEqual; &Greater Equal; 0.058 0.058 \end{matrix}$

其中：I—生育期总天数。 Among them: I—total days of reproductive period.

第i-1天结束时，第i天的实测土壤水分初始值θ_i即为已知；若第i天预测的土壤水分初始值θ_i'与实测土壤水分初始值θ_i非常接近，第i-1天的作物系数就取初始值；如果θ_i'与θ_i相差比较大，由实测土壤水分值θ_i可反推第i-1天的作物实际需水量： At the end of day i-1, the measured initial value of soil moisture θ _i on day _i is known; if the initial value of soil moisture θ _i ' predicted on day i is very close to the measured initial value of soil moisture The crop coefficient of -1 day takes the initial value; if the difference between θ _i ' and θ _i is relatively large, the actual water demand of the crop on the i-1 day can be deduced from the measured soil moisture value θ _i :

ET′_i-1＝1000nH_i-1θ_i-1+P_0i-1+W′_ri+M_i-1-1000nH_iθ_i ET′ _i-1 ＝1000nH _i-1 θ _i-1 +P _0i-1 +W′ _ri +M _i-1 -1000nH _i θ _i

其中：ET_i-1′—修正后的第i-1天作物的实际需水； Among them: ET _i-1 ′—the actual water requirement of crops on day i-1 after correction;

M_i-1—第i-1天的灌水量； M _i-1 — irrigation amount on day i-1;

则修正后第i-1天的作物实际需水量为： Then the actual water requirement of crops on day i-1 after correction is:

ET′_i-1＝W_i-1+P_0i-1+W′_ri+M_i-1-W_i' ET′ _i-1 ＝W _i-1 +P _0i-1 +W′ _ri +M _i-1 -W _i '

式中：W_i-1、W_i′——分别为第i-1天的初始、结束时的土壤含水量； In the formula: W _i-1 , W _i ′——soil water content at the beginning and end of day i-1, respectively;

进而可得到修正后的第i-1天作物系数K′_ci-1。 Then the corrected crop coefficient K′ _ci-1 on day i-1 can be obtained.

K'_c,i-1＝ET′_i-1/(K'_w,i-1·ET_0,i-1) K' _c,i-1 ＝ET' _i-1 /(K' _w,i-1 · ET _0,i-1 )

即在第i-1天结束时修正了当天的作物系数值，并把该修正值作为下一个计算时段的输入值，依此类推，实现对全生育期作物系数K_c值的逐日修正。 That is, the crop coefficient value of the day is corrected at the end of the i-1 day, and the corrected value is used as the input value of the next calculation period, and so on, so as to realize the daily correction of the crop coefficient K _c value in the whole growth period.

进一步，还包括对作物计划湿润层深逐日初始值确定模型： Further, it also includes a model for determining the daily initial value of the crop plan wetting layer depth:

${H h}_{i i} = = {h h}_{n no - - 11} + + (({h h}_{n no} - - {h h}_{n no - - 11})) \cdot \cdot ((i i - - {Σ Σ}_{j j = = 11}^{n no} {l l}_{j j - - 11}^{j j - - 11}_{j j - - 11})) / / {l l}_{n no}^{n no}$

其中：H_i——作物第i天(阶段)的计划湿润层深度； Among them: H _i ——planned wetting layer depth on the i-th day (stage) of the crop;

h_n-1——第n个生育期初始时计划湿润层深度； h _n-1 ——planned wetting layer depth at the beginning of the nth growth period;

h_n——第n个生育期结束时计划湿润层深度； h _n ——planned wetting layer depth at the end of the nth growth period;

n——作物所处生育期； n - the growth period of the crop;

i——作物播种后的生长累积天数； i - cumulative days of crop growth after sowing;

——第n个生育期的生长天数； - the number of days of growth in the nth growth period;

——第j个生育期的生长天数，j＝1,2,…,n。 ——the growth days of the jth growth period, j=1,2,...,n.

本土壤水分的分层测定及土壤水分的实时修正。本系统将作物根系活动区域以上土层视为一个整体系统，从所监测土壤水分数据中选取当天2时、8时、14时、20时等四个监测时刻的实测土壤水含水率，取作物所处计划湿润层深度以上所有传感器不同时刻所测土壤含水率的均值做为当天的土壤水分实测值。 The layered determination of soil moisture and the real-time correction of soil moisture. This system regards the soil layer above the root activity area of crops as a whole system, selects the measured soil water moisture content at four monitoring times of 2:00, 8:00, 14:00, and 20:00 from the monitored soil moisture data, and takes the crop The average value of the soil moisture content measured by all sensors above the depth of the planned wet layer at different times is taken as the measured value of the soil moisture of the day.

令θw_i(j，h_l)表示第i天在j监测时刻位于土层深h_l处传感器所测的土壤含水量，则第i天2时、8时、14时、20时在土层深h_l处的土壤水分值就可以分别记为：θw_i(2，h_l)、θw_i(8，h_l)、θw_i(14，h_l)、θw_i(20，h_l)，则可采用算数平均法计算第i天给定土层深h_l处的土壤水分值 Let θw _i (j, h _l ) represent the soil water content measured by the sensor at the depth h _l of the soil layer at the monitoring time j on the i-th day, then at 2 o'clock, 8 o'clock, 14 o'clock, The soil moisture values at depth h _l can be recorded as: θw _i (2, h _l ), θw _i (8, h _l ), θw _i (14, h _l ), θw _i (20, h _l ) , then the soil moisture value at a given soil depth h _l on the i-th day can be calculated using the arithmetic mean method

${θ θ}_{i i,, {h h}_{l l}} = = \frac{{θw θw}_{i i} ((22,, {h h}_{l l})) + + {θw θw}_{i i} ((88,, {h h}_{l l})) + + {θw θw}_{i i} ((1414,, {h h}_{l l})) + + θ θ {w w}_{i i} ((2020,, {h h}_{l l}))}{44}$

采用加权均值法则可得到第i天计划湿润层的土壤水分值θ_i： The soil moisture value θ _i of the planned wetting layer on the i-th day can be obtained by using the weighted mean method:

${θ θ}_{i i} = = {Σ Σ}_{l l = = 11}^{m m} {w w}_{i i} {θ θ}_{i i,, {h h}_{l l}}$

式中：w_i—为土壤层深度h_l的土壤水分值对θ_i的影响权重； In the formula: w _i — is the soil moisture value of the soil layer depth h _l The influence weight on θ _i ;

m—土层个数。 m—the number of soil layers. the

在进行土壤含水率预测时，以天为计算时段，递推过程如下：生育期第一次运行时，在生育期第一天开始时实测一次土壤含水率，作为阶段初始值，利用土壤水分逐日递推公式(6)逐日递推每一天阶段末的土壤含水率，并和当日实测的土壤含水率进行对比和修正，并把修正后的土壤含水率作为下一阶段的初始值，如此逐日顺序递推，进行每日计划湿润层含水率的模拟和修正。 When predicting soil moisture content, days are used as the calculation period, and the recursive process is as follows: when the growth period is running for the first time, the soil moisture content is measured once at the beginning of the first day of the growth period, and it is used as the initial value of the stage. Recursion formula (6) recursively calculates the soil moisture content at the end of each day, compares and corrects it with the actual measured soil moisture content of the day, and takes the corrected soil moisture content as the initial value of the next stage, so the order of the day Recursively, carry out the simulation and correction of the moisture content of the wet layer in the daily plan.

在进行计算的过程中，如果推算得到第i日含水率小于或者等于作物所处生育期最低允许含水率，同时考虑天气预报情况，在无雨或降雨量极少的情况下，采用第5.2节的实时灌溉预报模型对该日做出灌溉预报，并进行网络发布，当日土壤含水率修正到灌水后的土壤含水率，再以此为初始值进行计划湿润层含水率的递推，直至生育期结束。如果天气预报有降雨发生，则需要考虑有效降雨量，采用土壤水分逐日递推 $θ_{i} = \frac{H_{i - 1}}{H_{i}} θ_{i - 1} - ({ET}_{i - 1} - P_{0 i - 1} - W_{Ti} - M_{i - 1}) / (1000 n H_{i})$ 对计划湿润层的含水率进行分析预测，在作物生育期如有较大降雨发生，致使模拟计算过程中出现土壤含水率超过田间持水率的情况时，则将土壤含水率处理为田间持水率，多余的水量渗入深层土壤。 In the process of calculation, if the moisture content on the i-th day is estimated to be less than or equal to the minimum allowable moisture content in the growth period of the crop, and taking into account the weather forecast, in the case of no rain or very little rainfall, use Section 5.2 The real-time irrigation forecast model makes irrigation forecast for this day and publishes it on the network. The soil moisture content of the day is corrected to the soil moisture content after irrigation, and then the moisture content of the planned wet layer is deduced from this as the initial value until the growth period Finish. If there is rainfall in the weather forecast, it is necessary to consider the effective rainfall, and use soil moisture to deduce day by day $θ_{i} = \frac{h_{i - 1}}{h_{i}} θ_{i - 1} - ({ET}_{i - 1} - P_{0 i - 1} - W_{Ti} - m_{i - 1}) / (1000 no h_{i})$ Analyze and predict the moisture content of the planned wetting layer. If heavy rainfall occurs during the crop growth period, causing the soil moisture content to exceed the field water holding rate during the simulation calculation process, the soil moisture content will be treated as field water holding capacity rate, the excess water seeps into the deep soil.

作物系数K_ci初值的确定，在首次运行系统时，确定作物系数K_ci的初值，K_ci的计算初值采用随作物生育期累计天数i逐日变化的计算方法 The determination of the initial value of the crop coefficient K _ci is to determine the initial value of the crop coefficient K _ci when the system is run for the first time, and the initial value of K _ci is calculated according to the daily change of the cumulative days i of the crop growth period

${K K}_{ci ci} = = \{\begin{matrix} 7.346 7.346 {((i i / / I I))}^{22} - - 1.606 1.606 ((i i / / I I)) + + 0.0972 0.0972 & i i / / I I \leq \leq 0.058 0.058 \\ - - 3.463 3.463 ln ln ((i i / / I I)) - - 0.1909 0.1909 & i i / / I I &GreaterEqual; &Greater Equal; 0.058 0.058 \end{matrix}$

所述作物系数实时修正模型为K'_c,i-1＝ET′_i-1/(K'_w,i-1·ET_0,i-1)，式中，ET′_i-1为修正后第i-1天的作物实际需水量，计算公式为：ET′_i-1＝W_i-1+P_0i-1+W′_ri+M_i-1-W_i'，K'_w,i-1为第i-1天的实际土壤水分修正系数，W_i-1和W_i-1'分别为第i-1天的初始、结束时的土壤含水量，M_i-1为第i-1天的灌水量，P_0i-1为第i-1天的有效降雨量。 The real-time correction model of the crop coefficient is K' _c,i-1 = ET' _i-1 /(K' _w,i-1 · ET _0,i-1 ), where ET' _i-1 is the corrected The actual water requirement of crops on day i-1, the calculation formula is: ET′ _i-1 ＝W _i-1 +P _0i-1 +W′ _ri +M _i-1 -W _i '，K' _{w,i- 1} is the actual soil moisture correction coefficient on the i-1th day, W _i-1 and W _i-1 ' are the soil moisture content at the beginning and end of the i-1th day respectively, M _i-1 is the i-1th day The irrigation amount of the day, P _0i-1 is the effective rainfall of the i-1th day.

第i-1天结束时，第i天的实测土壤水分初始值θ_i即为已知；若第i天预测的土壤水分初始值θ_i'与实测土壤水分初始值θ_i非常接近，第i-1天的作物系数就取初始值；如果θ_i'与θ_i相差比较大，由实测土壤水分值θ_i可反推第i-1天的作物实际需水量： At the end of day i-1, the measured initial value of soil moisture θ _i on day _i is known; if the initial value of soil moisture θ _{i '} predicted on day i is very close to the measured initial value of soil moisture The crop coefficient of -1 day takes the initial value; if the difference between θ _i ' and θ _i is relatively large, the actual water demand of the crop on the i-1 day can be deduced from the measured soil moisture value θ _i :

M_i-1—第i-1天的灌水量； M _i-1 — irrigation amount on day i-1;

则修正后第_i-₁天的作物实际需水量为： Then the actual water requirement of crops on day _i - ₁ after correction is:

发明的系统功能包括： Invented system features include:

1、进行远程在线灌溉实时指导。该管理系统基于网络，根据短期天气预报结果和土壤的实时含水率结果，根据模型对数据采集系统实时采集的数据进行分析、归纳、处理等二次加工，计算作物的实时蒸腾蒸发量和在线进行作物实时非充分灌溉模拟、灌溉预报和灌溉管理，通过模型模拟、参数仿真等系统内部处理，精准确定作物是否需要灌溉、灌溉的水量和灌溉的时间等决策，为用户提供灌溉决策的实时咨询，并提供协助用水户确定不同灌溉方案的服务功能，对农业灌溉做出实时指导。能够实现作物灌溉的远程控制、信息管理和实时监控，还可以对数据进行备份、查询和统计，生成数据的统计报表。 1. Real-time guidance for remote online irrigation. The management system is based on the network, according to the short-term weather forecast results and the real-time soil moisture content results, and according to the model, the data collected in real time by the data acquisition system is analyzed, summarized, processed and other secondary processing is performed to calculate the real-time transpiration and evaporation of crops and online. Crop real-time insufficient irrigation simulation, irrigation forecast and irrigation management, through internal processing such as model simulation and parameter simulation, accurately determine whether crops need irrigation, irrigation water volume and irrigation time and other decisions, and provide users with real-time consultation on irrigation decisions, It also provides service functions to assist water users in determining different irrigation schemes, and provide real-time guidance on agricultural irrigation. It can realize remote control of crop irrigation, information management and real-time monitoring, and can also back up, query and count data, and generate statistical reports of data.

2、远程实时显示数据信息曲线图。远程实现农田微环境系统实时灌溉信息的采集、查询和存贮管理，动态显示土壤含水率等实时数据的变化情况，利用有效数据对监控对象的状况进行分析、管理，能够连续提供土壤墒情的实时变化，以及图形和数据前后处理。 2. Remote real-time display of data information graphs. Remotely realize the collection, query and storage management of real-time irrigation information of the farmland micro-environment system, dynamically display the changes of real-time data such as soil moisture content, use effective data to analyze and manage the status of monitoring objects, and continuously provide real-time information on soil moisture. changes, as well as pre- and post-processing of graphics and data.

3、实时进行作物生长期的生长模拟。利用模型库中的模型对作物根区土壤含水率进行实时模拟，对根区水分、作物的根深进行动态实时仿真，展示实时监控结果。 3. Real-time growth simulation of crop growth period. Use the models in the model library to simulate the soil moisture content in the root zone of crops in real time, perform dynamic real-time simulations on the water content in the root zone and the root depth of crops, and display the real-time monitoring results.

4、进行作物土壤水分的实时监控与在线传输。对作物根区土壤含水率进行分层测量和实时监测，完成系统数据的采集、远程传输和实时导入。 4. Real-time monitoring and online transmission of crop soil moisture. Carry out layered measurement and real-time monitoring of soil moisture content in the root zone of crops, and complete system data collection, remote transmission and real-time import.

附图说明 Description of drawings

图1是系统主要功能模块图； Figure 1 is a diagram of the main functional modules of the system;

图2是系统硬件测量及数据输出总体结构图； Figure 2 is an overall structure diagram of system hardware measurement and data output;

图3是农田实时灌溉系统数据库ER图； Fig. 3 is the ER diagram of the farmland real-time irrigation system database;

图4是采集实时墒情数据示意图； Fig. 4 is a schematic diagram of collecting real-time moisture data;

图5是作物实时灌溉模型程序编制流程图； Fig. 5 is a flow chart of crop real-time irrigation model programming;

图6是作物系数修正流程图； Fig. 6 is a flow chart of crop coefficient correction;

图7是实时灌溉中作物系数K_c具体修正图。 Fig. 7 is a specific correction diagram of the crop coefficient K _c in real-time irrigation.

具体实施方式 Detailed ways

实施例1：参见图1，田间智能灌溉在线控制管理方法包括田间试验站监控与管理单元和中心站管理单元，其中，田间试验站监控与管理单元包括气象数据采集模块和土壤墒情监测模块；中心站管理单元包括用户管理模块、数据上传模块、田间管理模块、农作物信息管理、作物灌溉模拟、农田实时监控、实时灌溉预报、图表显示、田间信息查询和信息统计。 Embodiment 1: Referring to Fig. 1, the field intelligent irrigation online control management method includes a field test station monitoring and management unit and a central station management unit, wherein the field test station monitoring and management unit includes a meteorological data acquisition module and a soil moisture monitoring module; the center The station management unit includes user management module, data upload module, field management module, crop information management, crop irrigation simulation, farmland real-time monitoring, real-time irrigation forecast, chart display, field information query and information statistics.

1用户管理模块 1 user management module

用户管理模块主要对系统用户进行管理，对系统用户可以进行查询、添加、修改和删除操作。用户通过用户名和密码可登陆系统，用户管理模块存储了用户的真实姓名、联系方式等信息，用户可查看、更改和添加自己的信息。 The user management module mainly manages system users, and can query, add, modify and delete system users. The user can log in to the system through the user name and password. The user management module stores the user's real name, contact information and other information. The user can view, change and add their own information.

系统可通过网络进行远程访问，为了保证系统的安全稳定，除了用户信息和联系方式外，模块还设置了用户的双权限管理模式，即用户根据权限分为普通用户身份和管理员身份。普通权限用户仅可以对系统内的显示数据和图像进行浏览访问，无法对系统数据进行添加、修改和删除等操作；而高级权限用户，即管理员用户，不仅可以浏览访问系统数据和图表，而且有权对系统的各种信息和数据进行管理和操作。 The system can be accessed remotely through the network. In order to ensure the security and stability of the system, in addition to user information and contact information, the module also sets up a dual-authority management mode for users, that is, users are divided into ordinary user identities and administrator identities according to their authority. Ordinary authority users can only browse and access the display data and images in the system, and cannot perform operations such as adding, modifying, and deleting system data; while advanced authority users, that is, administrator users, can not only browse and access system data and charts, but also Have the right to manage and operate various information and data in the system.

2数据上传模块 2 data upload module

数据上传模块主要是导入或导出数据库中的数据，对系统进行维护。 The data upload module mainly imports or exports the data in the database to maintain the system.

3田间管理模块 3 Field Management Module

田间管理模块中主要包含的信息为土壤参数值。在田间管理模块，可以实现对土壤信息、土壤水适宜含水率上、下限值的查询、添加、修改和删除功能，方便用户对多个田块土壤状况进行管理与监控。 The main information contained in the field management module is the soil parameter value. In the field management module, the functions of querying, adding, modifying and deleting the soil information and the upper and lower limits of the suitable moisture content of soil water can be realized, which is convenient for users to manage and monitor the soil conditions of multiple fields.

4农作物信息管理 4 Crop information management

农作物信息模块可以呈现出所监控的农作物信息，包括农作物的种类、田块、土壤含水率上、下限、雨量站、作物种植日期、作物收获日期、田间初始土壤水含水率等信息。用户可按照需求对以上信息进行查询、添加、修改和删除，并通过Java语言中先进的Ajax技术对土壤墒情信息、降雨信息以及相应的监测数据等进行同步更新和异步调取，为作物模拟和灌溉预报做充分准备。 The crop information module can display the monitored crop information, including crop types, fields, upper and lower limits of soil moisture content, rainfall stations, crop planting dates, crop harvest dates, and initial soil moisture content in the field. Users can query, add, modify and delete the above information according to their needs, and use the advanced Ajax technology in the Java language to update and asynchronously update and asynchronously retrieve soil moisture information, rainfall information and corresponding monitoring data. Prepare for irrigation forecasts.

5作物灌溉模拟 5 crop irrigation simulation

作物灌溉模拟模块功能可以根据数据库中的数据，利用模型库中的模型对作物生长和根区土壤含水率、作物系数等参数实现逐日递推，制定作物的实时灌溉制度，并采用实测结果对模拟结果进行实时修正，从而对模拟模型中的参数进行校验和修正，确定作物系数、根深等参数与作物生长时数的依变关系。 According to the data in the database, the function of the crop irrigation simulation module can use the model in the model library to realize the daily recursion of crop growth, root zone soil moisture content, crop coefficient and other parameters, formulate the real-time irrigation system of crops, and use the measured results to simulate The results are corrected in real time, so that the parameters in the simulation model are verified and corrected, and the dependence relationship between crop coefficient, root depth and other parameters and crop growth hours is determined.

登陆成功后可以实现以下几部分功能： After successful login, the following functions can be realized:

计算作物根区土壤含水率：根据需要从种植日期计算某一段时间内的土壤日含水率均值； Calculate the soil moisture content in the root zone of the crop: calculate the average daily soil moisture content within a certain period of time from the planting date as required;

模拟灌溉日期与灌溉量：根据降雨信息、气象站的风速、气压等信息，以及系统配置参数，模拟计算面临阶段否需要灌溉、何时灌溉、灌多少等问题； Simulate irrigation date and irrigation amount: According to rainfall information, wind speed, air pressure and other information of weather stations, as well as system configuration parameters, the simulation calculation faces problems such as whether irrigation is needed, when irrigation is needed, and how much irrigation is needed;

进行土壤水分修正系数k_w的修正：根据实测的土壤含水率，对模拟计算结果进行修正。 Correction of the soil moisture correction coefficient k _w : Correct the simulation calculation results according to the measured soil moisture content.

6农田实时监控 6 Farmland real-time monitoring

系统对监控对象进行远程监控，利用有效数据对监控对象进行分析和管理，并利用曲线图的直观性对监控对象进行动态实时仿真，展示实时监控结果。 The system remotely monitors the monitored objects, uses effective data to analyze and manage the monitored objects, and uses the intuitiveness of the graph to perform dynamic real-time simulation on the monitored objects to display real-time monitoring results.

7实时灌溉预报 7Real-time irrigation forecast

基于非充分灌溉理论、实时在线非充分实时灌溉预报和模拟模型，利用短期的气象预报和田间土壤水分实时监测数据，充分考虑降水的利用，进行灌溉制度的1日、3日和7日不同时段的预报。通过系统模型运转，可以远程实时确定作物的灌溉时间、灌溉定额，同时实现对作物系数K_c的实时修正。 Based on the theory of insufficient irrigation, real-time online insufficient real-time irrigation forecast and simulation model, using short-term weather forecast and real-time monitoring data of field soil moisture, fully considering the use of precipitation, the different periods of the irrigation system on the 1st, 3rd and 7th forecast. Through the operation of the system model, the irrigation time and irrigation quota of crops can be determined remotely and in real time, and the real-time correction of crop coefficient _Kc can be realized at the same time.

此模块主要实现对灌溉日期的预测。该模块由管理员自当前阶段，顺序向后推算1日、3日、7日，由作物灌溉制度模拟模型，可得到阶段末的土壤含水量，与设定的土壤含水率下限值进行比较，从而判断面临阶段是否需要灌溉，以及灌水日期。 This module mainly realizes the prediction of irrigation date. In this module, the administrator calculates the 1st, 3rd, and 7th days backwards from the current stage. The crop irrigation system simulation model can obtain the soil moisture content at the end of the stage, and compare it with the set lower limit value of soil moisture content. , so as to judge whether irrigation is needed in the facing stage, and the date of irrigation.

8图表显示 8 chart display

可实现对降雨、蒸发、土壤湿润层、土壤时段含水率、土壤平均含水率、以及灌溉数据的查询、添加、修改和删除功能，对作物系数经验值以及作物系数模拟值进行比较，并可对统计的数据以图表的形式进行直观的展示。 It can realize the functions of querying, adding, modifying and deleting rainfall, evaporation, soil wet layer, soil moisture content during time period, average soil moisture content, and irrigation data, compare the crop coefficient experience value and crop coefficient simulation value, and can Statistical data is visually displayed in the form of charts.

9田间信息查询 9Field Information Query

可实现对田间信息的单一或组合查询。用户根据土壤名称、信息时间、数据值进行单一或组合查询，为用户提供了多种查询方式，更加方便用户对田间信息进行查询。 Single or combined query of field information can be realized. Users can perform single or combined query based on soil name, information time, and data value, providing users with multiple query methods, making it more convenient for users to query field information.

10信息统计 10 Information statistics

实现土壤含水量相关的信息统计。当选定了某一田块后，即可直观的显示查询总灌溉次数、总灌水量、次灌溉量以及灌溉时间等。 Realize information statistics related to soil moisture content. When a field is selected, it can visually display and query the total irrigation times, total irrigation water volume, secondary irrigation volume and irrigation time, etc.

系统的软硬件组成 System hardware and software composition

参见图2系统硬件测量及数据输出总体结构图，采用温度传感器、湿度传感器、降水检测仪器、风速风向检测仪器和光照强度传感器等监测大气环境，利用土壤水分采集器(如EnviroScan)对根系不同深度的土壤水分进行检测，以数据库作为中间层，运用接口对数据的传输，完成对土壤状况的实时采集、处理和监控。 Refer to Figure 2 for the overall structure of system hardware measurement and data output. Temperature sensors, humidity sensors, precipitation detection instruments, wind speed and direction detection instruments, and light intensity sensors are used to monitor the atmospheric environment. It uses the database as the middle layer and uses the interface to transmit data to complete the real-time collection, processing and monitoring of soil conditions.

本发明以实现农田灌溉的自动化、智能化管理为目标，以土壤水环境实时监测设备EnviroScan和农田小型自动气象站AWS为硬件依托，以气象资料和农田基本信息为基础，针对已有作物实时灌溉预报中未能充分利用降雨的不足，以归纳形成的数学模型为依据，对土壤水分进行预测和分析，辅助决策者进行决策，用以实现农田水分灌溉自动化的智能管理系统。针对系统的总体设计，综合管理系统数据库ER图如图3所示。这一系统的应用与发展将实现作物水分管理的精准、高效定量决策，为节水农业发展提供一条新的解决方法和思路。 The invention aims to realize the automatic and intelligent management of farmland irrigation, relies on the real-time soil water environment monitoring equipment EnviroScan and the farmland small-scale automatic weather station AWS as the hardware, and based on the meteorological data and basic farmland information, real-time irrigation for existing crops The lack of rainfall is not fully utilized in the forecast. Based on the mathematical model formed by induction, the soil moisture is predicted and analyzed to assist decision makers in making decisions, so as to realize the intelligent management system of farmland water irrigation automation. For the overall design of the system, the ER diagram of the integrated management system database is shown in Figure 3. The application and development of this system will realize accurate and efficient quantitative decision-making of crop water management, and provide a new solution and idea for the development of water-saving agriculture.

与国内其他类似系统的区别及优越性如下： The differences and advantages from other similar systems in China are as follows:

1、能够实现灌溉管理的信息采集－处理－决策－信息反馈－监控为一体的节水实时在线综合管理。针对目前我国尚缺乏在线灌溉管理系统的问题，以作物非充分在线实时灌溉模型和预报模型为核心理论，运用面向对象思想，采用JAVA语言，开发了可视化、交互式的农田灌溉在线实时综合管理系统，实现了信息采集－处理－决策－信息反馈－监控为一体的节水综合管理。该系统以充分利用降雨，实时、适量的水量分配为出发点，实现了对农田实时灌溉的网络化和远程化管理。系统可以快速反应出参数的变化，与自动气象站和土壤水环境实时监测系统EnviroScan所测的数据同步；能够依据天气变化和作物生长状况做出灌溉决策和预报建议，系统对提高农业用水效率，节约农业水资源，缓解农业用水矛盾，解决半旱地农业粮食安全生产问题，实现农田灌溉的现代化具有重要的现实意义和巨大的社会、经济效益。 1. It can realize water-saving real-time online comprehensive management integrating information collection-processing-decision-making-information feedback-monitoring of irrigation management. Aiming at the current lack of online irrigation management system in our country, we developed a visualized and interactive online real-time integrated management system for farmland irrigation with the help of object-oriented thinking and JAVA language, based on the core theory of crop insufficiency online real-time irrigation model and forecasting model. , Realized the integrated water-saving management of information collection-processing-decision-making-information feedback-monitoring. The system starts from making full use of rainfall and distributing real-time and appropriate water, and realizes networked and remote management of real-time irrigation of farmland. The system can quickly respond to changes in parameters and synchronize with the data measured by the automatic weather station and the real-time soil and water environment monitoring system EnviroScan; it can make irrigation decisions and forecast suggestions based on weather changes and crop growth conditions. The system can improve agricultural water use efficiency, Saving agricultural water resources, alleviating the contradiction of agricultural water use, solving the problem of agricultural food safety production in semi-arid land, and realizing the modernization of farmland irrigation have important practical significance and huge social and economic benefits.

2、在灌溉预报模型中首次采用实时的关键参数。针对现有国内外灌溉管理中，灌溉模型关键参数采用阶段均值导致模型失真的现状，提出了灌溉和预报模型中关键参数的确定与实时修正方法，田间智能灌溉在线控制管理方法采用的是短期实时参数，属国内外首创，实现了作物非充分在线实时灌溉模拟和预报。 2. For the first time, real-time key parameters are used in the irrigation forecast model. Aiming at the current situation that the key parameters of the irrigation model are distorted by using the stage mean value in the existing irrigation management at home and abroad, a method of determining and real-time correction of the key parameters in the irrigation and forecasting model is proposed. The online control management method of field intelligent irrigation adopts short-term real-time parameter, which is the first at home and abroad, and realizes the online real-time irrigation simulation and forecasting of insufficient crops.

3、灌溉预报：针对已有作物灌溉预报中未能充分利用降雨的不足，已有的系统缺少对作物实时根区数据和面临阶段天气情况的考虑，特别是在应用实施时缺乏可操作性等问题。该系统结合田间的测量硬件，得到农作物生长过程中周围环境对其生长状况影响因素的监测结果，调用系统的模型库，基于短期天气预报，可以预测未来时段(1日、3日、7日等)的土壤含水量，实时模拟作物生长过程中对水分需求的动态变化，及时掌握农作物的生长状况，并且根据作物根区的土壤含水量进行灌溉预报，从而达到充分利用降水，有效提高农业水资源利用率的效果。所制定的灌溉制度既符合作物的生长需要，能满足其对土壤水环境的要求，又最大限度地提高了资源的有效利用率，能够在保证质量、提高作物产量的同时降低生产成本。 3. Irrigation forecast: In view of the lack of full use of rainfall in the existing crop irrigation forecast, the existing system lacks consideration of the real-time root zone data of the crop and the weather conditions in the upcoming stage, especially the lack of operability in the application implementation, etc. question. The system combines the measurement hardware in the field to obtain the monitoring results of the factors affecting the growth of the crops in the surrounding environment during the growth process, calls the model library of the system, and based on the short-term weather forecast, can predict the future period (1st, 3rd, 7th, etc. ) soil water content, real-time simulation of the dynamic changes in water demand during crop growth, timely grasp the growth status of crops, and forecast irrigation according to the soil moisture content in the root zone of crops, so as to make full use of precipitation and effectively improve agricultural water resources The effect of utilization. The irrigation system formulated not only meets the growth needs of crops, but also meets the requirements of soil and water environment, and maximizes the effective utilization of resources. It can reduce production costs while ensuring quality and increasing crop yield.

4、实现节水的非充分实时灌溉：针对实时在线非充分灌溉制度和灌溉预报技术中未充分利用降水、模型关键参数采用阶段均值的不足，系统模型库建立了作物系数、根深和作物生长时数的依变关系，建立了以1日、3日和7日为时段的作物非充分在线实时灌溉模拟和灌溉预报模型，实现了充分利用降水、精准实时灌溉、节约农业用水、提高灌溉水利用率的目的。可应用于限水灌溉和充分灌溉两种情况，帮助用户利用各种灌溉策略进行限量和充分供水决策，缺水一旦达到某一限定程度，系统就会根据实际情况，在考虑灌溉系统基础设置的前提下进行灌溉安排。 4. Insufficient real-time irrigation to achieve water saving: In view of the insufficient utilization of precipitation in the real-time online insufficient irrigation system and irrigation forecasting technology, and the lack of using stage mean values for key parameters of the model, the system model library has established crop coefficients, root depths and crop growth time. According to the dependent relationship of data, an online real-time irrigation simulation and irrigation forecast model of insufficient crops with 1st, 3rd, and 7th periods was established, which realized the full use of precipitation, precise real-time irrigation, saving agricultural water, and improving irrigation water utilization. rate purposes. It can be applied to water-limited irrigation and full irrigation, and helps users use various irrigation strategies to make limited and sufficient water supply decisions. Once the water shortage reaches a certain limit, the system will consider the basic settings of the irrigation system according to the actual situation. Under the premise of irrigation arrangements.

5、精准、智能灌溉：系统通过大量的信息和相关数据的整理、分析来进行土壤墒情预报，制定相应最优的实时在线灌溉制度，对科学、精准灌溉决策提供技术支持，实现对农田水分灌溉的远程、实时、自动化、网络化的智能管理，方便管理人员和用水户对系统的了解、掌握和使用。 5. Accurate and intelligent irrigation: The system forecasts soil moisture through the collation and analysis of a large amount of information and related data, formulates the corresponding optimal real-time online irrigation system, provides technical support for scientific and precise irrigation decisions, and realizes irrigation of farmland water The remote, real-time, automated and networked intelligent management makes it convenient for managers and water users to understand, master and use the system.

6、快速准确地实现大量信息的存储、查询、处理和及时更新，避免可利用信息的丢失；同时，能够及时、直观、形象和全方位的展现复杂系统特征的信息，对作物进行实时的灌溉预报，并且展现结果多以图表、数据形式出现，以Internet为展示平台，易于用户掌握和理解。 6. Quickly and accurately realize the storage, query, processing and timely update of a large amount of information to avoid the loss of available information; at the same time, it can timely, intuitively, vividly and comprehensively display the information of complex system characteristics, and perform real-time irrigation on crops Forecast and display results are mostly in the form of charts and data, and the Internet is used as the display platform, which is easy for users to grasp and understand.

7、简易性和可扩展性：系统简单方便，用户只需通过网络、无需安装专门的客户端即可进入系统，操作方便；系统可实现功能的不断完善与调整，能够根据用户的需求方便扩展，灵活控制不同区域、不同研究对象的灌水决策。 7. Simplicity and scalability: The system is simple and convenient. Users only need to enter the system through the network without installing a special client. It is easy to operate; the system can realize continuous improvement and adjustment of functions, and can be easily expanded according to user needs , flexibly control irrigation decisions in different areas and different research objects.

数据的实时采集、获取与处理： Real-time collection, acquisition and processing of data:

实时采集的内容包括气象数据采集模块和土壤墒情监测模块。 The content of real-time collection includes meteorological data collection module and soil moisture monitoring module.

1气象资料观测 1 Meteorological data observation

在田间设置有小型自动气象站(AWS)，用以测量每天的气象信息，测量内容包括2m风速、风向、太阳辐射、土壤温度、雨量、空气湿度、空气温度、大气压等共计10项。各项气象数据可采用人工手动的方式获取，下载到用户的笔记本电脑，也可以采用无线传输的方式直接获取数据。 A small automatic weather station (AWS) is set up in the field to measure daily meteorological information. The measurement content includes 2m wind speed, wind direction, solar radiation, soil temperature, rainfall, air humidity, air temperature, atmospheric pressure, etc., a total of 10 items. Various meteorological data can be obtained manually, downloaded to the user's laptop, or directly obtained by wireless transmission.

2土壤水分测定及获取 2 Soil moisture measurement and acquisition

土壤水分是作物生长的重要生态因素，土壤水分的准确测定是实行实时灌溉的一项重要依据。本系统采用目前最先进的土壤水环境监测设备EnviroScan来实时测定作物根区的土壤含水率。 Soil moisture is an important ecological factor for crop growth, and accurate measurement of soil moisture is an important basis for real-time irrigation. This system uses the most advanced soil water environment monitoring equipment EnviroScan to measure the soil moisture content in the root zone of crops in real time.

使用时系统时，要求在田间按预设位置埋设EnviroScan探测器，每组EnviroScan探测器共安装有16个传感器，测量0-2m土层内的土壤水分数值。每一组探测器包括2个EnviroScan探测器，分别监测0-1.2m、1.2m-2.0m的作物根区土壤水分，两个探测器间的埋设位置间隔为2m。考虑到北方代表性作物-冬小麦的根系生长特点，在0-1.2m作物根区，探测器上每隔10cm安装一个传感器；在1.2-2.0m作物根区，探测器上每隔20cm安装一个传感器。探测器通过电缆与RT6数据采集器相连，每个EnviroScan探测器对应一个RT6数据采集器。数据可以以 sdb格式或者excel格式输出，且两格式间数据可自由转换，采集时间可根据用户需要任意设置，传感器采集的土壤墒情数据以占土壤体积百分比计(见图4)。数据可以以excel形式直接下载使用，或者存储到系统实时数据库供模型调用。监测数据初步分析。为了解作物整个生育期间的土壤水分动态变化情况，对田间作物生育期土壤水分动态变化进行初步分析。 When using the system, it is required to bury EnviroScan detectors at preset positions in the field. Each group of EnviroScan detectors is equipped with a total of 16 sensors to measure the soil moisture value in the 0-2m soil layer. Each group of detectors includes 2 EnviroScan detectors, which monitor the soil moisture in the root zone of crops at 0-1.2m and 1.2m-2.0m respectively, and the buried position interval between the two detectors is 2m. Considering the root growth characteristics of winter wheat, a representative crop in the north, in the 0-1.2m crop root zone, install a sensor every 10cm on the detector; in the 1.2-2.0m crop root zone, install a sensor every 20cm on the detector . The detectors are connected to RT6 data collectors through cables, and each EnviroScan detector corresponds to an RT6 data collector. The data can be output in sdb format or excel format, and the data between the two formats can be freely converted. The collection time can be set arbitrarily according to the user's needs. The soil moisture data collected by the sensor is calculated as a percentage of the soil volume (see Figure 4). The data can be directly downloaded and used in the form of excel, or stored in the real-time database of the system for model calling. Preliminary analysis of monitoring data. In order to understand the dynamic changes of soil moisture during the whole growth period of crops, a preliminary analysis was made on the dynamic changes of soil moisture during the growth period of crops in the field.

田间智能灌溉模拟与灌溉预报(国内的灌溉模型在进行实时灌溉预报时，多是在雨后对土壤水分初始值进行修正，缺乏考虑灌前降雨预报的影响，特别是缺乏可用于实际操作的实时灌溉模型和实时灌溉预报技术)： Field intelligent irrigation simulation and irrigation forecast (domestic irrigation models mostly correct the initial value of soil moisture after the rain when performing real-time irrigation forecast, and lack of consideration of the impact of rainfall forecast before irrigation, especially the lack of real-time data that can be used in actual operations. irrigation models and real-time irrigation forecasting techniques):

1实时灌溉预报 1 Real-time irrigation forecast

在线实时灌溉预报是以农田“实时”资料包括土壤基本参数和观测的气象资料为基础，根据最新作物生长动态，实现对作物需水状况的准确、实时预估，并结合田间墒情预测及天气预报情况，及时做出灌溉决策，包括灌水日期和灌水定额等。实时灌溉预报是编制动态用水计划的基础。传统的作物灌溉预报过程中，常常不考虑未来短期内的气象变化，致使灌溉预报结果与实际生产需求不相符合，如可能会做出在降水前需要灌溉等不合理的灌水预报，从而造成降水无法充分利用，灌溉水资源也不能发挥其较高的生产效率，这对于水资源紧缺地区来讲，是对水资源的一种极大浪费。 The online real-time irrigation forecast is based on the "real-time" data of the farmland, including basic soil parameters and observed meteorological data. According to the latest crop growth dynamics, the accurate and real-time prediction of crop water demand is realized, combined with field moisture forecast and weather forecast. According to the situation, make timely irrigation decisions, including irrigation dates and irrigation quotas. Real-time irrigation forecasts are the basis for dynamic water use planning. In the process of traditional crop irrigation forecasting, meteorological changes in the short term in the future are often not considered, resulting in irrigation forecast results that do not match actual production needs. For example, unreasonable irrigation forecasts such as irrigation before precipitation may be made, resulting in precipitation. It cannot be fully utilized, and irrigation water resources cannot exert their high production efficiency, which is a great waste of water resources for areas where water resources are scarce.

为尽可能的利用降水资源，缓解缺水地区的水资源紧缺状况，进行实时灌溉预报时，不仅要逐日循序地进行作物的水量平衡计算，而且当作物根区土壤含水率降至作物可忍耐的土壤含水量下限附近，在做灌溉预报时就要进一步考虑未来短时期内可能出现的气象情况，根据来水情况预估作物的实际需水状况，并针对不同情形给出相应的解决办法，进而制定相应的灌水计划。 In order to make the best use of precipitation resources and alleviate the water shortage in water-scarce areas, when performing real-time irrigation forecasts, it is necessary not only to calculate the water balance of crops sequentially on a daily basis, but also to Near the lower limit of soil water content, when making irrigation forecasts, it is necessary to further consider the meteorological conditions that may occur in a short period of time in the future, estimate the actual water demand of crops according to the incoming water conditions, and provide corresponding solutions for different situations, and then Make a corresponding irrigation plan.

在实时灌溉预报时，首先需要根据地形条件、土壤质地、作物品种、作物生育阶段、农田小气候等条件，进行初始状态的修正，确定阶段参数的初始值；接着对所有影响田间水量平衡要素和影响灌水日期、灌水定额等因素进行逐日递推和分析，结合未来田间信息预测值和气象预报情况，做出灌溉决策。同时利用实测信息对相关模型参数进行逐时段修正与自修正，促使作物实时灌溉模型中的参数逐渐趋于稳定，不断提高灌溉预报结果的准确性和实时性。 In the real-time irrigation forecast, it is first necessary to correct the initial state and determine the initial value of the stage parameters according to the terrain conditions, soil texture, crop species, crop growth stage, farmland microclimate and other conditions; Irrigation date, irrigation quota and other factors are recursively and analyzed day by day, and irrigation decisions are made in combination with the predicted value of future field information and weather forecast. At the same time, the measured information is used to correct and self-correct the relevant model parameters step by step, so that the parameters in the crop real-time irrigation model gradually become stable, and the accuracy and real-time performance of the irrigation forecast results are continuously improved.

因此，农田在线实时灌溉中，如何利用作物需水模型准确的模拟作物生长过程中的需水动态，是在线实时灌溉预报的核心，如何建立土壤墒情预测模型科学的预测土壤水分动态，是灌溉预报的基础。在作物灌溉预报模型中，往往涉及到多个参数如计划湿润层的深度、作物系数、土壤水分修正系数等，参数的精确程度也是影响模拟结果的重要因素。而充分利用实时监测信息和反馈信息对作物需水模型、土壤墒情模型中的参数进行逐日的修正与自修正成为提高模型精度的有效方法，也是实时灌溉预报研究的重点和关键。 Therefore, in online real-time irrigation of farmland, how to use the crop water demand model to accurately simulate the dynamics of water demand during crop growth is the core of online real-time irrigation forecasting. How to establish a soil moisture prediction model to scientifically predict soil moisture dynamics is an important aspect of irrigation forecasting Foundation. In the crop irrigation forecasting model, multiple parameters are often involved, such as the depth of the planned wet layer, crop coefficient, soil moisture correction coefficient, etc. The accuracy of the parameters is also an important factor affecting the simulation results. Making full use of real-time monitoring information and feedback information to correct and self-correct the parameters in the crop water demand model and soil moisture model every day has become an effective method to improve the accuracy of the model, and it is also the focus and key of real-time irrigation forecast research.

2作物非充分在线实时灌溉模型 2 Crop Insufficient Online Real-time Irrigation Model

作物在线实时灌溉模型是在短期气象预报的基础上，根据灌区各田块实际土壤水分、作物实际蒸腾蒸发量、作物生长状况等，通过建立土壤墒情预测模型、作物需水预测模型、实时灌溉预报模型、作物系数实时修正模型等，实现实时预测作物在面临阶段是否需要灌水、灌水量以及灌水时间等。研究在对土壤水分实时监测的基础上，通过作物非充分在线实时灌溉模型，充分考虑面临阶段的降雨情况，尽可能准确、实时地预测未来不同时段的土壤水分动态含量，进而制定合理的灌溉方案，为农业生产提供在线指导和决策参考，为节约农业用水、提高灌溉水利用率提供科学依据。 The crop online real-time irrigation model is based on the short-term weather forecast, according to the actual soil moisture of each field in the irrigated area, the actual transpiration and evaporation of the crop, and the growth status of the crop, etc., by establishing a soil moisture prediction model, a crop water demand prediction model, and a real-time irrigation forecast. Model, crop coefficient real-time correction model, etc., to realize real-time prediction of whether crops need irrigation, irrigation amount and irrigation time in the facing stage. On the basis of real-time monitoring of soil moisture, the research uses the incomplete online real-time irrigation model of crops, fully considers the rainfall situation in the facing stage, predicts the dynamic content of soil moisture in different periods of time in the future as accurately and in real time as possible, and then formulates a reasonable irrigation plan , provide online guidance and decision-making reference for agricultural production, and provide scientific basis for saving agricultural water and improving irrigation water utilization.

(1)土壤墒情预测模型 (1) Soil moisture prediction model

土壤墒情是指一定体积土壤中所含水分的状况，墒情预测即根据前期测墒结果，结合气象条件，通过一定的手段来预测未来某一时期土壤含水量的多少。土壤墒情预测是灌溉预报的基础，对在水资源短缺条件下所进行的农田水分的合理调控具有重要意义。 Soil moisture refers to the moisture content in a certain volume of soil. Moisture prediction is to predict the amount of soil moisture in a certain period in the future based on the results of moisture measurement in the previous period and combined with meteorological conditions. Soil moisture forecasting is the basis of irrigation forecasting, and it is of great significance to the rational regulation of farmland moisture under the condition of water shortage.

本系统主要运用水量平衡原理分析制定旱作物的灌溉制度，建立土壤水量平衡模型进行土壤墒情预报。以天为预报时段的作物计划湿润层的水量平衡模型如下： This system mainly uses the principle of water balance to analyze and formulate the irrigation system of dry crops, and establishes a soil water balance model to forecast soil moisture. The water balance model of the crop plan wetting layer with days as the forecast period is as follows:

W_i＝W_i-1+P_0i+W_Ti-ET_i+M_i+K_i (1)式中：W_i-1——第i天初始计划湿润层的土壤含水量，mm； W _i ＝W _i-1 +P _0i +W _Ti -ET _i +M _i +K _i (1) In the formula: W _i-1 ——the soil water content of the initial planned wet layer on the i-th day, mm;

W_i——第i天结束时计划湿润层的土壤含水量，mm； W _i —soil moisture content of the planned wetting layer at the end of the i-th day, mm;

P_0i——第i天的有效降雨量，mm； P _0i - the effective rainfall on the i-th day, mm;

W_Ti——第i天由计划湿润层增加而增加的水量，mm； W _Ti ——the amount of water increased by the increase of the planned wetting layer on the i-th day, mm;

ET_i——第i天的作物需水量，mm； ET _i ——crop water requirement on day i, mm;

M_i——第i天的灌水量，mm； M _i ——Irrigation amount on the i-th day, mm;

K_i——第i天的地下水补给量，mm。 K _i — groundwater recharge on the i-th day, mm.

若研究区域的地下水埋深较深，地下水对作物的补给可以忽略不计，故式(1)可变为： If the buried depth of groundwater in the study area is relatively deep, the supply of groundwater to crops can be ignored, so formula (1) can be transformed into:

W_i＝W_i-1+P_0i+W_Ti-ET_i+M_i (2) W _i ＝W _i-1 +P _0i +W _Ti -ET _i +M _i (2)

其中，第i天初始、结束时计划湿润层土壤含水量可分别由下列各式求出： Among them, the soil water content of the planned wetting layer at the beginning and end of the i-th day can be calculated by the following formulas:

W_i-1＝1000nH_i-1θ_i-1 (3) W _i-1 = 1000nH _i-1 θ _i-1 (3)

W_i＝1000nH_iθ_i (4) W _i ＝1000nH _i θ _i (4)

式中：H_i-1—第i天初始计划湿润层深，mm； In the formula: H _i-1 — initial plan wet layer depth on the i-th day, mm;

H_i—第i天结束时的计划湿润层深，mm； H _i —planned wetting layer depth at the end of day i, mm;

θ_i-1—第i天初始土壤含水率，以占土壤体积百分比计，％； θ _i-1 —Initial soil moisture content on the i-th day, in percent of soil volume, %;

θ_i—第i天结束时的土壤含水率，以占土壤体积百分比计，％； θ _i —soil moisture content at the end of day i, in percentage of soil volume, %;

n—土壤孔隙率，％。 n—soil porosity, %.

由公式(5)可得第i时段内因计划湿润层的增加而增加的水量W_Ti。 The amount of water W _Ti increased due to the increase of the planned wetting layer within the i-th period can be obtained from formula (5).

W_Ti＝1000(H_i-H_i-1)·n·θ_deep (5) W _Ti ＝1000(H _i -H _i-1 )·n·θ _deep (5)

式中：θ_deep—深层土壤含水率，％； In the formula: θ _deep — deep soil moisture content, %;

由公式(2)、(3)、(4)可得出作物计划湿润层含水率的逐日递推预测模型： From the formulas (2), (3) and (4), the daily recursive prediction model of the moisture content of the crop plan wetting layer can be obtained:

${θ θ}_{i i} = = \frac{{H h}_{i i - - 11}}{{H h}_{i i}} {θ θ}_{i i - - 11} - - (({ET ET}_{i i - - 11} - - {P P}_{00 i i - - 11} - - {W W}_{Ti Ti} - - {M m}_{i i - - 11})) / / ((10001000 n no {H h}_{i i})) - - - - - - ((66))$

(2)作物实时需水量的模糊聚类预测模型 (2) Fuzzy clustering prediction model of crop real-time water demand

作物在线实时灌溉模型中，作物实时需水量ET_i的计算对作物的灌溉影响最大，是进行灌溉预报的关键。 In the crop online real-time irrigation model, the calculation of crop real-time water demand ET _i has the greatest impact on crop irrigation and is the key to irrigation forecast.

作物的实际需水量ET受到作物品种、气象条件、土壤类型、种植条件等诸多因素的影响，变化规律十分复杂。在计算非充分灌溉条件下作物需水量时，除去作物自身的因素，还必须考虑土壤水分低于适宜含水量下限时对蒸发蒸腾的抑制作用。在分析作物需水量计算方法的基础上，综合考虑作物及土壤水分的影响，本系统中分别构建了以1日、3日和7日为计算时段的作物的实时需水量计算模型： The actual water demand ET of crops is affected by many factors such as crop varieties, meteorological conditions, soil types, planting conditions, etc., and its changing rules are very complicated. When calculating the water requirement of crops under the condition of insufficient irrigation, the factors of the crop itself must be removed, and the inhibition of evapotranspiration when the soil moisture is lower than the lower limit of the suitable water content must also be considered. On the basis of analyzing the calculation method of crop water demand, and comprehensively considering the influence of crops and soil moisture, the real-time water demand calculation model of crops with 1 day, 3 days and 7 days as the calculation period is respectively constructed in this system:

ET_i＝K_ci·K_wi·ET_0i (7) ET _i =K _ci ·K _wi ·ET _0i (7)

式中：ET_i—第i天(阶段)的作物需水量，mm； In the formula: ET _i — crop water requirement on the i-th day (stage), mm;

ET_0i—第i天(阶段)的参考作物需水量，mm； ET _0i — reference crop water requirement on day i (stage), mm;

K_ci—第i天(阶段)的作物系数； K _ci —the crop coefficient of the i-th day (stage);

K_wi—实行非充分灌溉时第i天(阶段)的土壤水分修正系数； _Kwi — the soil moisture correction coefficient of the i-th day (stage) when non-full irrigation is implemented;

①参考作物需水量ET₀的模糊聚类预测 ① Fuzzy clustering prediction of reference crop water demand ET ₀

参考作物蒸发腾发量ET₀反映了大气环境对作物生长需水的影响程度，短期甚至逐日参考作物蒸发腾发量预测的准确性直接影响着实时灌溉的预报精度，是实时灌溉预报的重点和难点之一。参考作物需水量的计算方法有很多，常用的有：蒸发皿法、修正Penman公式法、Penman—Monteith法、Hargreaves方法、Priestley-Taylor方法等。当已获取实际的气象资料时，用修正的彭曼公式计算ET₀是最理想的。但在预测作物蒸发蒸腾量时，由于气象资料不可能完全准确预知，所以在实时预报中要依据不同的天气预测类型进行预报日ET₀的预估。 The reference crop evapotranspiration ET ₀ reflects the degree of influence of the atmospheric environment on the water demand for crop growth. The accuracy of the short-term or even daily reference crop evapotranspiration prediction directly affects the forecast accuracy of real-time irrigation, which is the focus and focus of real-time irrigation forecasting. One of the difficulties. There are many methods for calculating the water requirement of reference crops, and the commonly used ones are: evaporating pan method, modified Penman formula method, Penman-Monteith method, Hargreaves method, Priestley-Taylor method, etc. When the actual meteorological data have been obtained, it is ideal to calculate ET ₀ with the modified Penman formula. However, when predicting crop evapotranspiration, since meteorological data cannot be predicted completely and accurately, it is necessary to estimate the forecast day ET ₀ according to different weather forecast types in real-time forecasting.

根据实时灌溉预报的需要，研究将根据历史资料计算出的ET₀数据分别按照晴、云、阴、雨4种天气类型进行逐旬的分类统计，做成不同天气类型下的ET₀分类统计表。 According to the needs of real-time irrigation forecast, the research will calculate the ET ₀ data calculated according to the historical data according to the four weather types of sunny, cloud, cloudy and rain, and make the classification statistics of ET ₀ under different weather types. .

基于可变模糊集合理论对不同天气类型下的ET₀进行模糊聚类，得到不同天气类型的聚类中心s，并通过式(8)确定预报日天气类型下参考作物蒸发腾发量ET₀'值，从而进行ET₀的自修正： Based on the variable fuzzy set theory, fuzzy clustering is carried out on ET ₀ under different weather types to obtain the cluster centers s of different weather types, and the reference crop evapotranspiration ET ₀ ' under the forecast day weather type is determined by formula (8) value, thereby performing self-correction of ET ₀ :

ET′₀＝s·ET₀ (8) ET′ ₀ ＝s·ET ₀ (8)

其中，聚类中心s可以利用天气类型对作物潜在腾发ET₀模糊聚类分析的方法确定。以下是天气类型对作物潜在腾发ET₀的相对模糊聚类中心s算法。 Among them, the cluster center s can be determined by using the method of fuzzy cluster analysis of weather types to crop potential development ET ₀ . The following is the relative fuzzy clustering center s algorithm of weather type to crop potential evacuation ET ₀ .

设x_ij为不同天气类型下的ET₀值，其中，i表示天气类型，i＝1，2，3，4，分别表示晴、云、阴、雨；j表示旬，j＝1，2，……，36。由ET₀分类统计表可得全年各旬不同天气类型下多年平均ET₀的特征值矩阵： Let x _ij be the ET ₀ value under different weather types, wherein, i represents the weather type, i=1, 2, 3, 4, respectively represent sunny, cloud, overcast, rain; j represents ten days, j=1, 2, ..., 36. From the ET ₀ classification statistical table, the eigenvalue matrix of the multi-year average ET ₀ under different weather types in each ten-day of the year can be obtained:

${x x}_{ij ij} = = [\begin{matrix} {x x}_{1111} & {x x}_{1212} & . . . . . . & {x x}_{11 j j} \\ {x x}_{21 twenty one} & {x x}_{21 twenty one} & . . . . . . & {x x}_{22 j j} \\ . . . . . . & . . . . . . & . . . . . . & . . . . . . \\ {x x}_{i i 11} & {x x}_{i i 22} & . . . . . . & {x x}_{ij ij} \end{matrix}] - - - - - - ((99))$

式中：maxx_ij——j旬ET₀特征值的最大值。 In the formula: maxx _ij —— the maximum value of the eigenvalues of ET ₀ in days j.

$R R = = {r r}_{ij ij} = = [\begin{matrix} {r r}_{1111} & {r r}_{1212} & . . . . . . & {r r}_{11 j j} \\ {r r}_{21 twenty one} & {r r}_{22 twenty two} & . . . . . . & {r r}_{22 j j} \\ . . . . . . & . . . . . . & . . . . . . & . . . . . . \\ {r r}_{i i 11} & {r r}_{i i 22} & . . . . . . & {r r}_{ij ij} \end{matrix}] - - - - - - ((1111))$

$S S = = (({S S}_{ih i h})) = = [\begin{matrix} {S S}_{1111} & {S S}_{1212} \\ {S S}_{21 twenty one} & {S S}_{22 twenty two} \\ . . . . . . & . . . . . . \\ {S S}_{i i 11} & {S S}_{i i 22} \end{matrix}] - - - - - - ((1212))$

式中：s_ih——为分类模式h下i种天气类型下ET₀的特征值规格化数，0≤s_ih≤1； In the formula: s _ih —— is the normalized number of eigenvalues of ET ₀ under i weather types under the classification mode h, 0≤s _ih ≤1;

h＝1，2，…，c。 h=1, 2, . . . , c.

$U u = = (({u u}_{hj hj})) = = [\begin{matrix} {u u}_{1,1 1,1} & {u u}_{1,2 1,2} & . . . . . . & {u u}_{1,36 1,36} \\ {u u}_{2,1 2,1} & {u u}_{2,1 2,1} & . . . . . . & {u u}_{2,36 2,36} \end{matrix}] - - - - - - ((1313))$

且满足条件： And meet the conditions:

$\begin{matrix} 00 \leq \leq {u u}_{hj hj} \leq \leq 11 \\ {Σ Σ}_{h h = = 11}^{c c} {u u}_{hj hj} = = 00 \\ {Σ Σ}_{j j = = 11}^{n no} {u u}_{hj hj} > > 00 \end{matrix}\} - - - - - - ((1414))$

$| | | | {w w}_{i i} (({r r}_{j j} - - {s the s}_{h h})) | | | | = = {{{Σ Σ}_{i i = = 11}^{m m} [[{w w}_{i i} {(({r r}_{ij ij} - - {s the s}_{ih i h}))}^{22}]]}}^{\frac{11}{22}} - - - - - - ((1515))$

式中：i——天气类型总数。 In the formula: i—the total number of weather types.

F(u_hj)＝u_hj||w_i(r_j-s_h)|| (16) F(u _hj )＝u _hj ||w _i (r _j -s _h )|| (16)

①给定u_hj与s_ih所要求满足的计算精度ε₁、ε₂； ①Given u _hj and s _ih required to meet the calculation accuracy ε ₁ , ε ₂ ;

②假设一个满足约束条件(14)且元素不全部相等的初始模糊聚类矩阵 ② Assume an initial fuzzy clustering matrix that satisfies the constraint condition (14) and the elements are not all equal

③代入对应的初始模糊聚类中心矩阵 ③ substitute The corresponding initial fuzzy clustering center matrix

${S S}_{ih i h}^{00} = = \frac{{Σ Σ}_{j j = = 11}^{n no} {u u}_{hj hj}^{00^{22}} {w w}_{i i}^{22} {r r}_{ij ij}}{{Σ Σ}_{j j = = 11}^{n no} {u u}_{hj hj}^{00^{22}} {w w}_{i i}^{22}} - - - - - - ((1717))$

④代入求一次近似模糊聚类矩阵 ④ substitute Find an approximate fuzzy clustering matrix

${u u}_{hj hj}^{11} = = \frac{11}{{Σ Σ}_{h h = = 11}^{c c} \frac{{Σ Σ}_{i i = = 11}^{m m} {[[{w w}_{i i} - - (({r r}_{ij ij} - - {s the s}_{ih i h}^{00}))]]}^{22}}{{Σ Σ}_{i i = = 11}^{m m} {[[{w w}_{i i} - - (({r r}_{ij ij} - - {s the s}_{ih i h}))]]}^{22}}} - - - - - - ((1818))$

⑤代入求一次近似模糊聚类中心矩阵 ⑤ substitute Finding an approximate fuzzy clustering center matrix

⑥逐个比较矩阵和以及矩阵的对应元素，若且则迭代结束，为最优聚类中心矩阵s，否则重复(3)至(5)的步骤直至满足精度要求，已有研究已证明了迭代的收敛性。 ⑥ Comparing matrices one by one and and the matrix The corresponding elements of , if and then the iteration ends, is the optimal cluster center matrix s, otherwise Repeat the steps from (3) to (5) until the accuracy requirements are met, and previous studies have proved the convergence of iterations.

经模糊聚类计算，可以得到云、阴、雨天气情况下ET₀对于晴天ET₀的模糊聚类中心。二预报精度分析 Through fuzzy clustering calculation, the fuzzy clustering center of ET ₀ in cloudy, cloudy and rainy weathers to ET ₀ in sunny days can be obtained. 2 Forecast accuracy analysis

采用绝对累计偏差(ABE)、均方根误差(RMSE)、相对误差(RE)、决定系数(R²)和认同系数(IA)等指标来评定预报精度，各统计参数的评价标准见表1，统计变量具体表达式如式(19)～(25)。 Absolute cumulative deviation (ABE), root mean square error (RMSE), relative error (RE), coefficient of determination (R ² ) and identification coefficient (IA) are used to evaluate the forecast accuracy. The evaluation criteria of each statistical parameter are shown in Table 1 , the specific expressions of statistical variables are as formulas (19)~(25).

$ABE ABE = = {Σ Σ}_{k k = = 11}^{n no} | | {x x}_{k k} - - {y the y}_{k k} | | - - - - - - ((1919))$

$RMSE RMSE = = \sqrt{\frac{{Σ Σ}_{k k = = 11}^{m m} {(({x x}_{k k} - - {y the y}_{k k}))}^{22}}{n no}} - - - - - - ((2020))$

$RE RE = = \frac{RMSE RMSE}{\overset{&OverBar; &OverBar;}{x x}} - - - - - - ((21 twenty one))$

${R R}^{22} = = \frac{{[[Σ Σ (({x x}_{k k} - - \overset{&OverBar; &OverBar;}{x x})) (({y the y}_{k k} - - \overset{&OverBar; &OverBar;}{y the y}))]]}^{22}}{Σ Σ {(({x x}_{k k} - - \overset{&OverBar; &OverBar;}{x x}))}^{22} Σ Σ {(({y the y}_{k k} - - \overset{&OverBar; &OverBar;}{y the y}))}^{22}} - - - - - - ((22 twenty two))$

$IA IA = = 11 - - \frac{{Σ Σ}_{k k = = 11}^{n no} {(({x x}_{k k} - - {y the y}_{k k}))}^{22}}{{Σ Σ}_{k k = = 11}^{n no} {((| | {x x}_{k k}^{' '} | | + + | | {y the y}_{k k}^{' '} | |))}^{22}} - - - - - - ((23 twenty three))$

${x x}_{k k}^{' '} = = {x x}_{k k} - - \overset{&OverBar; &OverBar;}{x x} - - - - - - ((24 twenty four))$

${y the y}_{k k}^{' '} = = {y the y}_{k k} - - \overset{&OverBar; &OverBar;}{y the y} - - - - - - ((2525))$

表1预测结果精度评价标准 Table 1 Prediction results accuracy evaluation criteria

作物需水量ET的预测 Forecast of Crop Water Requirement ET

通过以上模糊聚类计算，即可确定不同天气类型对作物潜在腾发ET₀的模糊聚类中心矩阵s。根据式(8)完成对ET₀的自修正后，即可采用半经验公式(26)进行作物需水量ET的计算，即 Through the above fuzzy clustering calculation, the fuzzy clustering center matrix s of the potential development ET ₀ of different weather types to crops can be determined. After the self-correction of ET ₀ is completed according to formula (8), the semi-empirical formula (26) can be used to calculate crop water demand ET, namely

ET＝ET₀'·K_c·K_w＝s·ET₀·K_c·K_w (26)(3)有效降雨量P₀计算模型 ET＝ET ₀ '·K _c ·K _w ＝s·ET ₀ ·K _c ·K _w (26)(3) Calculation model of effective rainfall P ₀

降水贮存于作物根区后，可以有效地被作物蒸腾蒸发所利用，从而降低作物的灌溉需水量，因此对于缺水地区而言，充分利用降水，可以有效地缓解水资源的紧缺现状。发生降水时，当降水强度大于土壤的入渗能力，或者降水超过土壤贮水能力时，降水量中会有一部分以地表径流形式流走，或形成深层渗漏流出作物根区，从而不能被作物所利用。因此，只有有效降水量才能够补充作物的需水要求。对非充分灌溉来讲，有效降雨量P₀是制定作物灌溉制度、灌溉用水管理的一个重要影响因素。 After precipitation is stored in the root zone of crops, it can be effectively used by crop transpiration and evaporation, thereby reducing the irrigation water demand of crops. Therefore, for water-scarce areas, making full use of precipitation can effectively alleviate the shortage of water resources. When precipitation occurs, when the precipitation intensity is greater than the infiltration capacity of the soil, or the precipitation exceeds the soil water storage capacity, part of the precipitation will flow away in the form of surface runoff, or form deep seepage and flow out of the root zone of the crop, so that it cannot be absorbed by the crop. used. Therefore, only effective precipitation can supplement the water requirement of crops. For insufficient irrigation, the effective rainfall P ₀ is an important factor influencing crop irrigation system and irrigation water management.

影响有效降水的因素很多，因计算目的不同，确定有效降水的估算方法也不尽相同。影响有效降雨量的主要因素有降雨特性、土壤特性等，一般生产中采用经验的降雨有效利用系数法计算有效降雨量P_0i，见公式(25)。 There are many factors that affect effective precipitation, and the estimation methods for determining effective precipitation are also different due to different calculation purposes. The main factors affecting the effective rainfall are rainfall characteristics, soil characteristics, etc. In general production, the effective rainfall utilization coefficient method is used to calculate the effective rainfall P _0i , see formula (25).

P_0i＝αP_i (3-25)式中：P_i—实际降雨量(mm)； P _0i = αP _i (3-25) where: P _i —actual rainfall (mm);

α—降雨有效利用系数，α取值大小与降雨量的大小、降雨强度、降雨延续时间、土壤性质、地面覆盖以及地形因素有关。α取值见表2。 α—Rainfall effective utilization coefficient, the value of α is related to the amount of rainfall, rainfall intensity, duration of rainfall, soil properties, ground cover and topographical factors. See Table 2 for the value of α.

表2降雨有效利用系数α取值 Table 2 Rainfall Effective Utilization Coefficient α Values

P(mm) P(mm) <5mm <5mm 5～50mm 5～50mm >50mm >50mm α alpha 0 0 0.8～1.0 0.8～1.0 0.7～0.8 0.7～0.8

(4)在线实时灌溉预报模型 (4) Online real-time irrigation forecast model

作物在线实时灌溉预报模型是基于水量平衡原理，结合土壤水分墒情变化和作物需水动态，确定合理、科学的灌水定额和灌水时间。良好的作物实时灌溉预报需要最新的土壤水分等实测资料以及短期气象预报资料，来对每一个面临阶段的预测结果进行实时修正并进行下一阶段的相关参数与决策的预测，同时还需考虑不同田块土壤性状，依据不同土壤储蓄水分的特点，通过动态数值模拟进行灌溉预报。 The crop online real-time irrigation forecast model is based on the principle of water balance, combined with changes in soil moisture and crop water demand dynamics, to determine a reasonable and scientific irrigation quota and irrigation time. A good real-time crop irrigation forecast requires the latest measured data such as soil moisture and short-term weather forecast data to make real-time corrections to the forecast results of each facing stage and to predict the relevant parameters and decision-making in the next stage. At the same time, it is necessary to consider different Field soil properties, according to the characteristics of different soil water storage, irrigation forecast through dynamic numerical simulation.

根据水量平衡原理，建立以天为时段的作物在线实时灌溉预报模型： According to the principle of water balance, an online real-time irrigation forecast model for crops is established with the day as the time period:

M_i＝W_i-1+P_0i+W_Ti-ET_i+K_i-W_i (27) M _i =W _i-1 +P _0i +W _Ti -ET _i +K _i -W _i (27)

i一方面表示递推的第i个阶段，同时表示作物种植后的第i天，通过i将模型参数和作物生长相关联，以便寻求相互间的依变关系。 On the one hand, i represents the i-th stage of the recursion, and at the same time represents the i-th day after the crop is planted. Through i, the model parameters and crop growth are related, so as to seek the mutual dependence relationship.

式中各符号意义同以上各式。 The symbols in the formula have the same meaning as the above formulas.

为满足作物正常的生长发育需要，作物任意时段内计划湿润层的根系储水量必须经常保持在一定的适宜范围内。 In order to meet the normal growth and development needs of crops, the root water storage in the planned wetting layer of crops must always be kept within a certain suitable range in any period of time.

系统运行时，首先利用作物计划湿润层含水率的逐日递推预测模型对土壤墒情进行预测，当预测的土壤计划湿润层含水率θ_i不大于设定的土壤水分下限指标θ_c2时，对应的日期即为预测的灌水日期，需要进行灌溉预报和灌水量计算。由于在线实时灌溉模型受作物计划湿润层含水率的逐日递推预测模型中计划湿润层深H_i、作物系数等参数的影响，灌溉模型模拟结果会有一定的误差，因此在每个计算时段结束时，都需要用土壤含水率实测值对土壤含水率的模拟值进行误差分析与判断，同时对模型参数进行必要的修正。 When the system is running, first use the daily recursive forecasting model of the moisture content of the _crop planned wetting layer to predict the soil _moisture . The date is the predicted irrigation date, and irrigation forecast and irrigation volume calculation are required. Since the online real-time irrigation model is affected by parameters such as the planned wetting layer depth H _i and crop coefficient in the daily recursive prediction model of the moisture content of the crop planning wetting layer, the simulation results of the irrigation model will have certain errors, so at the end of each calculation period It is necessary to use the measured value of soil moisture content to analyze and judge the error of the simulated value of soil moisture content, and to make necessary corrections to the model parameters.

若预测时段的土壤计划湿润层含水率θ_i不大于设定的土壤水分下限指标θ_c2，但根据天气预报该时段内有降雨发生，则不灌水或推迟灌水。灌水后，以实测的土壤含水率作为下一阶段的初始土壤含水率，进行下一阶段的灌水日期和灌水量预报。如此进行逐阶段递推，从种植日开始直至作物生长期结束。 If the moisture content of the planned soil moisture layer θ _i during the forecast period is not greater than the set soil moisture lower limit index θ _c2 , but according to the weather forecast, there will be rainfall during this period, then irrigation will not be performed or irrigation will be postponed. After irrigation, the measured soil moisture content is used as the initial soil moisture content of the next stage to forecast the irrigation date and irrigation volume of the next stage. In this way, it is recursively carried out stage by stage, from the beginning of the planting day to the end of the crop growth period.

模型的计算机程序流程见图5。 The computer program flow of the model is shown in Figure 5.

(5)实时灌溉制度 (5) Real-time irrigation system

灌溉制度是否合理或高效主要取决于灌水时期与土壤水分状况变化和产量形成规律的吻合程度。 Whether the irrigation system is reasonable or efficient mainly depends on the degree of coincidence between the irrigation period and the change of soil moisture status and the law of yield formation.

(1)灌水定额 (1) Irrigation quota

灌水定额是单位面积上的一次灌水量，与作物种类、土壤持水量、灌溉面积以及可利用的灌水时间有关。灌水定额是灌溉制度的主要内容之一，确定合理的灌水定额，不仅可以正确指导农田灌溉节水高产，而且也是工程设计、水资源合理利用的主要依据。 Irrigation quota is the amount of irrigation water per unit area, which is related to crop type, soil water holding capacity, irrigated area and available irrigation time. Irrigation quota is one of the main contents of irrigation system. Determining a reasonable irrigation quota can not only correctly guide farmland irrigation to save water and increase yield, but also is the main basis for engineering design and rational use of water resources.

灌水定额通常按照计划湿润层内土壤含水率的上、下限及计划湿润层厚度计算。灌溉定额的计算如下： The irrigation quota is usually calculated according to the upper and lower limits of the soil moisture content in the planned wetting layer and the thickness of the planned wetting layer. The irrigation quota is calculated as follows:

来水充分时作物灌水量： Crop irrigation amount when water is sufficient:

当来水量很大，足以满足灌溉需求时，灌水量M_i为： When the amount of incoming water is large enough to meet the irrigation demand, the amount of irrigation water M _i is:

M_i＝1000·n·H_i(1-θ_i)·θ_max (28) M _i =1000·n·H _i (1-θ _i )·θ _max (28)

式中：θ_i—第i天(阶段)的初始土壤含水率； In the formula: θ _i — the initial soil moisture content on the i-th day (stage);

H_i—第i天(阶段)的作物计划湿润层深度，m； H _i —the depth of the crop plan wet layer on the i-th day (stage), m;

来水不充分时作物灌水量： Crop irrigation amount when water is not sufficient:

当来水不足或者水资源量紧缺时，对作物进行非充分灌溉。灌水量M_i为： Sub-irrigating crops when there is insufficient water or water resources are scarce. The irrigation volume M _i is:

M_i＝1000·n·H_i(θ_c1-θ_i)·θ_max (29) M _i =1000·n·H _i (θ _c1 -θ _i )·θ _max (29)

式中：θ_c1—灌溉后所要达到的土壤含水率，非充分灌溉时θ_c1一般取90％θ_max。 In the formula: θ _c1 — the soil moisture content to be achieved after irrigation, and θ _c1 is generally taken as 90% θ _max when not fully irrigated.

(2)灌水时间 (2) Irrigation time

由于农田土壤干湿状况会受到降水的强烈影响而发生波动，因此同一区域同一作物需水的关键时期在不同的水文年型也会发生变化。降水规律的不确定导致了作物根区干湿状况的不确定，尤其是在水资源紧缺的半干旱地区，更需要结合作物的耗水过程和产量形成过程深入研究，建立作物非充分在线实时灌溉制度，将有限的水量在作物的不同生育期进行合理分配。 Since the dryness and wetness of farmland soil will be strongly affected by precipitation and fluctuate, the critical period of water demand for the same crop in the same region will also change in different hydrological years. Uncertainty of precipitation law leads to uncertainty of dry and wet conditions of crop root zone, especially in semi-arid areas where water resources are scarce, it is necessary to combine in-depth research on crop water consumption process and yield formation process, and establish crop incomplete online real-time irrigation A system that rationally allocates the limited amount of water in different growth stages of crops.

实时灌溉预报是在实测气象资料的基础上，根据建立田间土壤水分的逐日递推模拟模型，对作物短期乃至逐日的土壤水分变化情况做出准确预报，当土壤含水率接近作物所处生育期最低允许含水率时，进行灌溉决策；如果预报期间有降水或灌溉补水发生，则推迟灌溉，并根据实际的具体情况对预报结果进行及时调整，同时修正土壤水分变化曲线。 The real-time irrigation forecast is based on the measured meteorological data, and according to the establishment of a daily recursive simulation model of soil moisture in the field, it can accurately predict the short-term and even daily soil moisture changes of crops. When the water content is allowed, the irrigation decision is made; if there is precipitation or irrigation supplementation during the forecast period, the irrigation is postponed, and the forecast results are adjusted in time according to the actual specific situation, and the soil moisture change curve is corrected at the same time.

3作物实时灌溉模型关键参数的确定及修正 3 Determination and correction of key parameters of crop real-time irrigation model

作物非充分在线实时灌溉模型，是在对土壤水分实现逐日监测的基础上，准确、实时地预测未来不同时段的土壤水分动态含量，并充分利用降雨预报信息，制定合理的灌溉方案，为发展节水农业、实现高效率的农田灌溉提供技术支撑。 The crop incomplete online real-time irrigation model is based on the daily monitoring of soil moisture, accurately and in real time predicting the dynamic content of soil moisture in different periods in the future, and making full use of the rainfall forecast information to formulate a reasonable irrigation plan for economic development. Provide technical support for water agriculture and high-efficiency farmland irrigation.

实时灌溉预报不仅需要最新的实测资料和预测资料，对每一阶段灌溉预测进行实时修正来进行下一阶段的预测，而且还需考虑不同田块的特点，对各田块土壤水分含量通过动态数值模拟进行灌溉预报。在实时灌溉模型中，涉及到土壤、作物、气象等各种参数，因此模型参数的确定与修正关系着灌溉预报结果的精确程度和实用程度。 Real-time irrigation forecasting not only requires the latest measured data and forecast data, but also corrects the irrigation forecast in each stage in real time to make the next stage forecast, and also needs to consider the characteristics of different fields, and calculate the soil moisture content of each field through dynamic values. Simulation for irrigation forecasting. In the real-time irrigation model, various parameters such as soil, crops, and weather are involved, so the determination and correction of the model parameters are related to the accuracy and practicality of the irrigation forecast results.

(1)适宜土壤水分下限指标的研究 (1) Research on the lower limit index of suitable soil moisture

土壤水分适宜下限值，是指适宜于作物生长的最低的土壤含水量指标。土壤含水率的大小与作物的生长有着密切的关系。当气候和土壤类型已定时，土壤含水率降到一定的范围，会对作物生长有限制作用。在进行非充分灌溉时，制定合理的土壤水分下限指标对于指导作物适时、适量的灌溉和节约水资源具有重要的意义。土壤水分的下限值决定作物的灌水时间，也影响作物灌水次数和灌水量的确定，可以通过制定适宜的水分下限来调控土壤水分，减少灌水量与灌水次数，进而提高作物水分利用效率。 The suitable lower limit of soil moisture refers to the lowest soil moisture index suitable for crop growth. Soil moisture content is closely related to the growth of crops. When the climate and soil type are determined, the soil moisture content drops to a certain range, which will limit the growth of crops. When under-irrigating, it is of great significance to establish a reasonable lower limit index of soil moisture to guide crops to irrigate in a timely and appropriate manner and save water resources. The lower limit of soil moisture determines the irrigation time of crops, and also affects the determination of the frequency and amount of crop irrigation. It is possible to regulate soil moisture by setting an appropriate lower limit of water, reduce the amount of irrigation and the frequency of irrigation, and then improve the water use efficiency of crops.

作物种类不同，对土壤水分下限的要求也不同。土壤水分对作物生长的作用随作物生长发育阶段的变化而变化。作物生育前期，土壤水分可促进营养生长，对苗数的多少和强弱起决定作用；作物生育中期，充足的水分会促进作物的生长发育，决定作物穗数的多少；生育后期，为保证作物正常灌浆充实，也必须有水分保证。同一作物的不同生育阶段对水分亏缺的敏感性也不同，敏感性越大，其缺水减产的损失越大。另外，土壤质地不同，其含水量及持水能力也不同，对土壤含水率下限的要求也不同。 Different types of crops have different requirements for the lower limit of soil moisture. The effect of soil moisture on crop growth varies with the stage of crop growth and development. In the early stage of crop growth, soil moisture can promote vegetative growth and play a decisive role in the number and strength of seedlings; in the middle stage of crop growth, sufficient water will promote the growth and development of crops and determine the number of ears of crops; Normal grouting must also have moisture guarantee. Different growth stages of the same crop have different sensitivities to water deficit, and the greater the sensitivity, the greater the loss of water shortage and reduced yield. In addition, the soil texture is different, its water content and water holding capacity are also different, and the requirements for the lower limit of soil moisture content are also different.

由于影响土壤水分下限指标的因素很多，不同作物、不同地区的土壤水分下限指标的计算方法及制定标准也有所不同。整体来说，大多是按作物生育阶段划分的，对短期内优化或者是按天来计算各参数的模型非常少，与实时灌溉结合不够密切，尚还需要这方面的研究。 Since there are many factors affecting the lower limit of soil moisture, the calculation methods and standards for the lower limit of soil moisture are different for different crops and regions. Generally speaking, most of them are divided according to the growth stage of crops. There are very few models for short-term optimization or calculation of parameters on a daily basis, and the combination with real-time irrigation is not close enough, and research in this area is still needed.

本系统中，先期利用SWAT模型中作物模块模拟了不同土壤水分上下限条件下冬小麦产量变化情况，通过方案比较，选取了冬小麦适宜的土壤水分上下限指标。 In this system, the crop module in the SWAT model was used to simulate the change of winter wheat yield under different upper and lower limits of soil moisture. Through the comparison of the schemes, the appropriate upper and lower limits of soil moisture for winter wheat were selected.

(2)计划湿润层深度初值的确定 (2) Determination of the initial value of the planned wet layer depth

计划湿润层深度是指需要通过灌溉补充土壤水分的土层深度。对旱作物来说，土壤计划湿润层深度通常是作物的主要根系吸水层，它主要取决于作物生长状况和作物根系活动层的深度，与作物品种、生育阶段、田间土壤性质、以及地下水埋深和土壤微生物活动等因素也有关系。 The planned wetting layer depth refers to the depth of the soil layer that needs to be replenished with soil moisture through irrigation. For dry crops, the depth of the soil plan wet layer is usually the main water-absorbing layer of the crop root system, which mainly depends on the growth status of the crop and the depth of the active layer of the crop root system, which is related to the crop variety, growth stage, field soil properties, and groundwater burial. Factors such as depth and soil microbial activity are also related.

计划湿润层深度直接影响着灌水定额的确定，其确定方法通常有两种，一种为动态型，即认为计划湿润层深度应随着作物的生长生育期、作物根系活动层深度以及地下水埋深等发生变化。如冬小麦在幼苗期计划湿润层深为0.3～0.4m，分蘖期为0.4～0.5m，拔节期为0.5～0.6m，抽穗期为0.6～0.8m，灌浆期为0.8～1.0m；玉米在幼苗期计划湿润层深0.3～0.4m，拔节期为0.4～0.5m，孕穗期为0.5～0.6m，抽穗期为0.6～0.8m，灌浆期为0.8；棉花在幼苗期计划湿润层深0.3～0.4m，现蕾期为0.4～0.6m，开花结铃期为0.6～0.8m，吐絮期为0.6～0.8m。春小麦在抽穗前计划湿润层按80cm来计算，在生育期其他时间的灌水计划湿润层按40～60cm考虑。 The depth of the planned wetting layer directly affects the determination of the irrigation quota. There are usually two ways to determine it. One is the dynamic type. Wait for changes. For example, the depth of the wet layer of winter wheat is 0.3-0.4m at the seedling stage, 0.4-0.5m at the tillering stage, 0.5-0.6m at the jointing stage, 0.6-0.8m at the heading stage, and 0.8-1.0m at the filling stage; The planned wetting layer depth is 0.3-0.4m at the early stage, 0.4-0.5m at the jointing stage, 0.5-0.6m at the booting stage, 0.6-0.8m at the heading stage, and 0.8m at the filling stage; the planned wetting layer depth at the seedling stage of cotton is 0.3-0.4m m, the budding stage is 0.4~0.6m, the flowering and boll setting stage is 0.6~0.8m, and the boll opening stage is 0.6~0.8m. The planned wetting layer of spring wheat before heading is calculated as 80 cm, and the planned wetting layer of irrigation at other times during the growth period is considered as 40-60 cm.

另一种则认为同一作物的计划湿润层深度在整个生育期应始终采用同一深度，但具体计划层深度在应用中各不相同。农作物的主要根系分布层是确定作物适宜湿润层深度的基本依据。 The other thinks that the planned wetting layer depth of the same crop should always be the same depth throughout the growth period, but the specific planned layer depth varies in application. The main root distribution layer of crops is the basic basis for determining the depth of the suitable wetting layer for crops.

本系统中，认为计划湿润层深度应随着作物的生长发育、根系的不断加深而增加。灌溉预报过程中，需要确定整个生育期第i天(阶段)的土壤计划湿润层深度初值，为此，假设作物全生育期各个生长阶段内，计划湿润层呈线性均匀增加，则作物任一天的计划湿润层深度可采用线性的逐日递推模型进行模拟，为此，建立作物计划湿润层深计算模型： In this system, it is considered that the depth of the planned wetting layer should increase with the growth and development of the crops and the continuous deepening of the root system. In the process of irrigation forecasting, it is necessary to determine the initial value of the depth of the soil planned wetting layer on the i-th day (stage) of the entire growth period. Therefore, assuming that the planned wetting layer increases linearly and uniformly in each growth stage of the whole growth period of the crop, the crop any day The planned wetting layer depth can be simulated using a linear day-by-day recursive model. Therefore, the calculation model of the crop planning wetting layer depth is established:

${H h}_{i i} = = {h h}_{n no - - 11} + + (({h h}_{n no} - - {h h}_{n no - - 11})) \cdot &Center Dot; ((i i - - {Σ Σ}_{j j = = 11}^{n no} {l l}_{j j - - 11}^{j j - - 11}_{j j - - 11})) / / {l l}_{n no}^{n no} - - - - - - ((3030))$

式中：H_i——作物第i天(阶段)的计划湿润层深度，m； In the formula: H _i ——the depth of the planned wetting layer on the i-th day (stage) of the crop, m;

h_n-1——第n个生育期初始时计划湿润层深度，m； h _n-1 ——planned wetting layer depth at the beginning of the nth growth period, m;

h_n——第n个生育期结束时计划湿润层深度，m； h _n ——planned wetting layer depth at the end of the nth growth period, m;

n——作物所处生育期； n - the growth period of the crop;

i——作物播种后的生长累积天数，d； i - cumulative days of crop growth after sowing, d;

——第n个生育期的生长天数，d； ——the growth days of the nth growth period, d;

——第j个生育期的生长天数，d，j＝1,2,…,n。 ——the growth days of the jth growth period, d, j=1,2,...,n.

确定出第i天的计划湿润层深初始值后，可通过Irrimax软件对土壤含水量监测值进行分析，得到实时的作物根深，进而对计划湿润层深的初始值进行修正，利用修正后的计划湿润层深，根据式(5)计算预报时段内由于计划湿润层增加而增加的水量W_Ti。 After determining the initial value of the planned wetting layer depth on the i-th day, the monitoring value of soil moisture content can be analyzed by Irrimax software to obtain the real-time crop root depth, and then the initial value of the planned wetting layer depth can be corrected, and the revised plan can be used Depth of wetting layer, according to formula (5), calculate the amount of water W _Ti increased due to the increase of planned wetting layer within the forecast period.

(3)土壤实时墒情监测与数据传输 (3) Real-time soil moisture monitoring and data transmission

土壤墒情是农田灌溉的一项重要指标，准确的农田墒情测报是实现作物适时适量灌溉的基础，是精准灌溉技术与灌区节水灌溉管理的依据。通过对农田土壤水分变化动态的监视，对灌溉作物需水量和有关参数进行分析和计算，能够实现对未来土壤墒情和旱情趋势的预报，从而制定准确的灌溉计划。 Soil moisture is an important indicator of farmland irrigation. Accurate forecasting of farmland moisture is the basis for timely and appropriate irrigation of crops, and the basis for precision irrigation technology and water-saving irrigation management in irrigation areas. By monitoring the dynamic changes of farmland soil moisture, analyzing and calculating the water demand of irrigated crops and related parameters, the forecast of future soil moisture and drought trends can be realized, so as to formulate accurate irrigation plans.

本系统将作物根系活动区域以上土层视为一个整体系统，从所监测土壤水分数据中选取当天2时、8时、14时、20时等四个监测时刻的实测土壤水含水率，取作物所处计划湿润层深度以上所有传感器不同时刻所测土壤含水率的均值做为当天的土壤水分实测值。 This system regards the soil layer above the root activity area of crops as a whole system, and selects the measured soil water moisture content at four monitoring times of 2:00, 8:00, 14:00, and 20:00 from the monitored soil moisture data, and takes the crop The average value of the soil moisture content measured by all sensors above the depth of the planned wetting layer at different times is taken as the actual measured value of soil moisture for the day.

设第i天作物所处计划湿润层深为100cm时，读取100cm以上(含100cm)不同土层处传感器的所测土壤水分值，记为θw_i(j，l)，其中j为观测时刻，l为传感器所处土层深，则第i天不同观测时刻不同土层深传感器所测土壤水分值可如图6表示。 Assuming that the planned wetting layer depth of the crop on the i-th day is 100cm, read the measured soil moisture values of the sensors at different soil layers above 100cm (including 100cm), denoted as θw _i (j, l), where j is the observed time, l is the depth of the soil layer where the sensor is located, then the soil moisture values measured by the sensors at different soil layer depths at different observation times on the i-th day can be shown in Figure 6.

${θ θ}_{i i,, {h h}_{l l}} = = \frac{{θw θw}_{i i} ((22,, {h h}_{l l})) + + {θw θw}_{i i} ((88,, {h h}_{l l})) + + {θw θw}_{i i} ((1414,, {h h}_{l l})) + + θ θ {w w}_{i i} ((2020,, {h h}_{l l}))}{44} - - - - - - ((3131))$

${θ θ}_{i i} = = {Σ Σ}_{l l = = 11}^{m m} {w w}_{i i} {θ θ}_{i i,, {h h}_{l l}} - - - - - - ((3232))$

m—土层个数。 m—the number of soil layers. the

在进行计算的过程中，如果推算得到第i日含水率小于或者等于作物所处生育期最低允许含水率，同时考虑天气预报情况，在无雨或降雨量极少的情况下，采用第5.2节的实时灌溉预报模型对该日做出灌溉预报，并进行网络发布，当日土壤含水率修正到灌水后的土壤含水率，再以此为初始值进行计划湿润层含水率的递推，直至生育期结束。如果天气预报有降雨发生，则需要考虑有效降雨量，采用土壤水分逐日递推公式(6)对计划湿润层的含水率进行分析预测，在作物生育期如有较大降雨发生，致使模拟计算过程中出现土壤含水率超过田间持水率的情况时，则将土壤含水率处理为田间持水率，多余的水量渗入深层土壤。 In the process of calculation, if the moisture content on the i-th day is estimated to be less than or equal to the minimum allowable moisture content in the growth period of the crop, and taking into account the weather forecast, in the case of no rain or very little rainfall, use Section 5.2 The real-time irrigation forecast model makes irrigation forecast for this day and publishes it on the network. The soil moisture content of the day is corrected to the soil moisture content after irrigation, and then the moisture content of the planned wet layer is deduced from this as the initial value until the growth period Finish. If there is rainfall in the weather forecast, it is necessary to consider the effective rainfall, and use the soil moisture daily recursion formula (6) to analyze and predict the moisture content of the planned wet layer. If there is a large rainfall during the crop growth period, the simulation calculation process When the soil water content exceeds the field water holding rate, the soil water content is treated as the field water holding rate, and the excess water infiltrates into the deep soil.

基于土壤墒情预测，采用作物非充分实时灌溉模型、实时灌溉预报模型则可进行作物的实时灌溉和预报。其余如3日、7日、及至2周的灌溉预报，计算的阶段步长相应分别为3日、7日和2周，原理同1日的灌溉预报。 Based on the prediction of soil moisture, real-time irrigation and forecasting of crops can be carried out by using crop insufficiency real-time irrigation model and real-time irrigation forecast model. For other irrigation forecasts such as 3 days, 7 days, and up to 2 weeks, the calculated step lengths are 3 days, 7 days, and 2 weeks respectively, and the principle is the same as that of the 1-day irrigation forecast.

(4)土壤水分修正系数K_wi的确定 (4) Determination of soil moisture correction coefficient K _wi

在非充分灌溉中，由于田间水分受到调节和控制，作物各生育阶段的根系生长、植株发育、群体结构以及生态指标均会受到影响。 Under insufficient irrigation, root growth, plant development, population structure, and ecological indicators of crops at each growth stage will be affected due to the regulation and control of field moisture.

对于旱作物，土壤中自凋萎点到田间持水率之间的水分，可以保持在根层内，并能被植物吸收利用。当土壤含水率在临界含水率与田间持水率之间时，土壤水分通过毛管作用充分供给作物蒸发、蒸腾的需要，土壤含水率的高低并不影响作物的蒸发蒸腾；当土壤含水率小于临界含水率时，土壤水分运移由于受到阻力的作用，运移的实际速率将会小于充分蒸发蒸腾时所需的速率，此时土壤含水率的高低就直接影响到蒸发蒸腾的速率。 For dry crops, the water in the soil from the withering point to the field water holding capacity can be kept in the root layer and can be absorbed and utilized by plants. When the soil moisture content is between the critical moisture content and the field water holding rate, the soil moisture can fully supply the needs of crop evaporation and transpiration through capillary action, and the level of soil moisture content does not affect the crop evapotranspiration; when the soil moisture content is less than At the critical moisture content, the actual rate of soil moisture migration will be lower than the rate required for full evapotranspiration due to the effect of resistance. At this time, the level of soil moisture content directly affects the rate of evapotranspiration.

因此，非充分灌溉条件下引进了土壤水分修正系数K_w来反映土壤水分不充足时土壤含水率对作物需水量的影响。当进行充分灌溉，土壤水分不限制作物蒸发蒸腾时，土壤水分修正系数k_w取值为1。 Therefore, under the condition of insufficient irrigation, the soil moisture correction coefficient _Kw was introduced to reflect the influence of soil moisture content on crop water demand when soil moisture is insufficient. When sufficient irrigation is carried out and soil moisture does not limit crop evapotranspiration, the soil moisture correction coefficient k _w takes a value of 1.

由于气象、土壤类型、作物种类、根系深度等因素的影响，k_w与θ的关系是非常复杂的。合理地选取以及确定作物系数和土壤水分修正系数对于准确地计算非充分灌溉条件下的作物蒸发蒸腾量有很大影响，对于非充分灌溉应用于生产实际也具有非常重要的指导意义。 Due to the influence of weather, soil type, crop species, root depth and other factors, the relationship between _kw and θ is very complicated. Reasonable selection and determination of crop coefficient and soil moisture correction coefficient have a great influence on the accurate calculation of crop evapotranspiration under insufficient irrigation conditions, and also have very important guiding significance for the application of insufficient irrigation in actual production.

结合研究区域实际情况，采用以下公式计算土壤水分修正系数K_wi： Combined with the actual situation in the study area, the soil moisture correction coefficient K _wi is calculated using the following formula:

${K K}_{wi wi} = = \{\begin{matrix} ln ln ((11 + + 100100 \frac{{θ θ}_{i i}}{{θ θ}_{max max}})) / / ln ln 101101 & {θ θ}_{c c 22} \leq \leq {θ θ}_{i i} < < {θ θ}_{c c 11} \\ α α \cdot \cdot exp exp [[(({θ θ}_{i i} - - {θ θ}_{c c 22}))]] / / {θ θ}_{c c 22} & {θ θ}_{i i} < < {θ θ}_{c c 22} \end{matrix} - - - - - - ((3333))$

式中：θ_i—第i天(阶段)土壤含水率，以占土壤体积百分比计，％； In the formula: θ _i —soil moisture content on the i-th day (stage), in percent of soil volume, %;

θ_max—田间持水率，以占土壤体积百分比计，％； θ _max —Field water holding capacity, in percentage of soil volume, %;

α—经验系数，旱作物可取0.89。 α—experience coefficient, 0.89 is desirable for dry crops.

4作物系数K_c的实时模拟与修正 4 Real-time simulation and correction of crop coefficient _Kc

作物系数K_c是用于估算作物需水量的重要参数，而作物实时需水规律及需水量的预测又是实时灌溉预报的核心，因此，准确的确定及模拟作物系数K_c对于农田发展精准灌溉、节约水资源具有重大意义。 The crop coefficient _Kc is an important parameter for estimating the water demand of crops, and the prediction of crop real-time water demand and water demand is the core of real-time irrigation forecasting. Therefore, the accurate determination and simulation of crop coefficient _Kc is very important for the development of precision irrigation in farmland. , Saving water resources is of great significance.

(1)作物系数K_ci初值的确定 (1) Determination of initial value of crop coefficient K _ci

在首次运行系统时，需要确定作物系数K_ci的初值。综合考虑实验区的气候条件、作物类型、模型精度等因素，K_ci的计算初值采用随作物生育期累计天数i逐日变化的计算方法： When running the system for the first time, the initial value of the crop coefficient K _ci needs to be determined. Considering the climatic conditions in the experimental area, crop types, model accuracy and other factors, the initial value of K _ci is calculated using the calculation method that changes day by day with the cumulative days i of the crop growth period:

${K K}_{ci ci} = = \{\begin{matrix} 7.346 7.346 {((i i / / I I))}^{22} - - 1.606 1.606 ((i i / / I I)) + + 0.0972 0.0972 & i i / / I I \leq \leq 0.058 0.058 \\ - - 3.463 3.463 ln ln ((i i / / I I)) - - 0.1909 0.1909 & i i / / I I &GreaterEqual; &Greater Equal; 0.058 0.058 \end{matrix} - - - - - - ((3434))$

式中：I—生育期总天数，d。 In the formula: I—total days of growth period, d.

(2)作物系数的K_ci修正 (2) K _ci correction of crop coefficient

以往的研究中，在利用作物系数K_c计算作物需水量时，多是采用按生育阶段划分K_c的平均值，缺乏以天为单位的作物系数值和对作物系数的实时、逐日修正，阶段均值的使用与实际多变的天气、土壤及作物生长状况显然不相符，从而致使作物的实时需水量计算和农田实时灌溉预报不够准确，影响作物的精准灌溉与有限农业水资源的高效利用，导致了作物灌溉预报的失真，失去了应有的指导生产的作用。 In previous studies, when using the crop coefficient _Kc to calculate crop water demand, the average value of _Kc divided by growth stage was mostly used, and there was a lack of crop coefficient values in days and real-time and daily corrections to crop coefficients. The use of the average value is obviously inconsistent with the actual changeable weather, soil and crop growth conditions, which leads to inaccurate real-time water demand calculation of crops and real-time irrigation forecast of farmland, which affects the precise irrigation of crops and the efficient use of limited agricultural water resources, resulting in The distortion of crop irrigation forecast has been lost, and the role of guiding production has been lost.

针对已有作物需水规律研究方面的不足，特别是需水模型中作物系数K_c常取生长阶段均值的问题，本研究充分利用灌溉监测实验与采集的实时信息成果，以及田间的气象资料，提出了作物系数的逐日预测与自修正方法，对作物系数K_c进行率定及逐步逼近修正，以更精准的模拟作物的实时需水规律，为作物在线灌溉制度研究提供基础科学信息。具体计算过程如下： In view of the deficiencies in the research on the law of crop water demand, especially the problem that the crop coefficient _Kc in the water demand model often takes the average value of the growth stage, this study makes full use of the real-time information results collected from irrigation monitoring experiments and field meteorological data. The daily prediction and self-correction method of crop coefficient is proposed, and the crop coefficient K _c is calibrated and gradually approximated to correct, so as to more accurately simulate the real-time water demand law of crops, and provide basic scientific information for the research of crop online irrigation system. The specific calculation process is as follows:

首先利用自动气象站实测气象资料计算出每天的参考作物需水量ET_0i，再由计算得到的土壤水分修正系数K_wi、作物系数K_ci、参考作物需水量ET_0i，根据公式(7)计算作物的实际需水量。由作物在线实时灌溉模型可知，在作物生长过程中，第(i-1)天结束时，该天的土壤含水量、有效降雨量及灌溉水量均已知，由此可根据土壤水分修正系数K_w的计算公式(33)计算出该天的实际土壤水分修正系数K'_w,i-1；而初始作物系数K_c值则采用同一地区的经验值。至此，就可以利用公式(7)计算出作物的实际需水量ET_i。由第(i-1)天的实测土壤水分初始值θ_i-1，根据土壤水分逐日递推公式(6)就可推算出第i天的土壤水分初始值θ_i'。 Firstly, the daily reference crop water requirement ET _0i is calculated using the weather data measured by the automatic weather station, and then the calculated soil moisture correction coefficient K _wi , crop coefficient K _ci , and reference crop water requirement ET _0i are calculated according to formula (7). actual water demand. According to the crop online real-time irrigation model, during the crop growth process, at the end of the (i-1)th day, the soil moisture content, effective rainfall and irrigation water volume of the day are known, so the coefficient K can be corrected according to the soil moisture The calculation formula (33) of _w calculates the actual soil moisture correction coefficient K' _w,i-1 of the day; and the initial crop coefficient K _c value adopts the empirical value of the same area. So far, the actual water requirement ET _i of the crop can be calculated by using formula (7). From the measured initial soil moisture value θ _i-1 on the (i-1)th day, the soil moisture initial value θ _i ' on the i-th day can be calculated according to the soil moisture daily recursion formula (6).

第(i-1)天结束时，第i天的实测土壤水分初始值θ_i即为已知。若第i天预测的土壤水分初始值θ_i'与实测土壤水分初始值θ_i非常接近，第(i-1)天的作物系数就取初始值；如果θ_i'与θ_i相差比较大，则由实测土壤水分值θ_i可反推第(i-1)天的作物实际需水量： At the end of the (i-1)th day, the measured initial value of soil moisture _θi on the i-th day is known. If the predicted initial value of soil moisture θ _i ' on the i-th day is very close to the measured initial value of soil moisture θ _i , the crop coefficient on the (i-1) day will take the initial value; if the difference between θ _i ' and θ _i is relatively large, Then the actual water demand of the crop on the (i-1) day can be deduced from the measured soil moisture value θ _i :

ET_i'_-1＝1000nH_i-1θ_i-1+P_0i-1+W_r'_i+M_i-1-1000nH_iθ_i (35)式中：ET_i-1′—修正后的第(i-1)天作物的实际需水量，mm； ET _i ' _-1 ＝1000nH _i-1 θ _i-1 +P _0i-1 +W _r ' _i +M _i-1 -1000nH _i θ _i (35) where: ET _i-1 ′—the modified first (i-1) The actual water demand of crops per day, mm;

M_i-1—第(i-1)天的灌水量，mm，可由公式(28)或(29)计算得到。 M _i-1 —Irrigation amount on the (i-1)th day, mm, can be calculated by formula (28) or (29).

则修正后第(i-1)天的作物实际需水量可写为： Then the actual water requirement of crops on day (i-1) after correction can be written as:

ET′_i-1＝W_i-1+P_0i-1+W′_ri+M_i-1-W_i' (36) ET' _i-1 ＝W _i-1 +P _0i-1 +W' _ri +M _i-1 -W _i ' (36)

式中：W_i-1、W_i′——分别为第(i-1)天的初始、结束时的土壤含水量，mm； In the formula: W _i-1 , W _i ′——respectively, the soil water content at the beginning and end of the (i-1)th day, mm;

进而可得到修正后的第(i-1)天作物系数K′_ci-1。 Then the corrected crop coefficient K' _ci-1 on the (i-1)th day can be obtained.

K'_c,i-1＝ET′_i-1/(K'_w,i-1·ET_0,i-1) (37) K' _c,i-1 ＝ET' _i-1 /(K' _w,i-1 · ET _0,i-1 ) (37)

即在第(i-1)天结束时修正了当天的作物系数值，并把该修正值作为下一个计算时段的输入值，依此类推，实现对全生育期作物系数K_c值的逐日修正。 That is, at the end of the (i-1) day, the crop coefficient value of the day is corrected, and this correction value is used as the input value of the next calculation period, and so on, so as to realize the daily correction of the crop coefficient K _c value in the whole growth period .

通过一个生育周期的修正，可得到逐日的作物系数。在地区和作物种类不变的情况下，将该组作物系数作为第二年在线实时灌溉模型中的作物系数初始值，然后进行作物的实时需水量计算，同时根据当年监测结果再次进行逐日修正。依此类推，对作物系数进行逐年自我修正，不断提高作物在线实时的灌溉预报精度，逐步实现农田精准灌溉。实时灌溉中作物系数K_c具体修正过程见图7。经过多年长期的自我修正，作物系数K_c值将逐步趋于稳定，接近于真实的K_c，灌溉预报也会更加精准。 Through a growth cycle correction, the daily crop coefficient can be obtained. Under the condition that the area and crop type remain unchanged, this group of crop coefficients is used as the initial value of the crop coefficient in the online real-time irrigation model in the second year, and then the real-time water demand of the crop is calculated, and the daily correction is performed again according to the monitoring results of the year. By analogy, the crop coefficient is self-corrected year by year, and the accuracy of crop online real-time irrigation forecast is continuously improved, and the precision irrigation of farmland is gradually realized. See Figure 7 for the specific correction process of the crop coefficient _Kc in real-time irrigation. After years of long-term self-correction, the crop coefficient K _c value will gradually become stable, close to the real K _c , and the irrigation forecast will be more accurate.

Claims

1. A field intelligent irrigation online control method, is characterized in that comprising the following steps:

Step 1, the small automatic weather station collects meteorological data, the soil water environment monitoring equipment measures the soil moisture content in the root zone of the crop in real time, and the collected and measured data are imported and stored in the database of the server located on the Internet;

Step 2, the client logs in to the farmland real-time intelligent irrigation software system located on the server;

Step 3, the farmland real-time intelligent irrigation software system invokes the data in the database, and performs field soil information management, crop information management, farmland real-time monitoring, field real-time irrigation simulation and irrigation forecast, field information query, information statistics.

2. The field intelligent irrigation online control method according to claim 1, characterized in that: the field real-time irrigation simulation and irrigation forecast, on the basis of short-term weather forecast, according to the actual soil moisture of each field in the irrigation area, the actual transpiration of crops Evaporation, crop growth status, through the establishment of soil moisture prediction model, crop water demand prediction model, real-time irrigation forecast model, crop coefficient real-time correction model, real-time prediction of whether crops need irrigation during the forecast period, irrigation time, irrigation amount, and in real time At the end of each calculation period, the simulated value of soil moisture content and crop coefficient are corrected with the measured value of soil moisture content.

3. A field intelligent irrigation online control management method, comprising a soil moisture detection unit for monitoring the moisture content W of the soil wet layer, a rainfall detection unit for monitoring the actual rainfall P, and an irrigation system, characterized in that it also includes Soil moisture forecasting model, fuzzy clustering forecasting model of crop real-time water demand and online real-time irrigation forecasting and control model;

The real-time prediction model of the soil moisture content: W _i =W _i-1 +P _0i +W _Ti -ET _i +M _i +K _i

Among them: W _i-1 ——the soil water content of the initial planned wet layer on the i-th day;

W _i —soil moisture content of the planned wetting layer at the end of the i-th day;

P _0i - the effective rainfall on the i-th day;

W _Ti ——the amount of water increased by the increase of the planned wetting layer on the i-th day;

ET _i ——crop water requirement on day i;

M _i ——Irrigation amount on the i-th day;

K _i — groundwater recharge on the i-th day;

The fuzzy clustering prediction model of the real-time water demand of the crops: ET=ET ₀ '·K _c ·K _w =s·ET ₀ ·K _c ·K _w ;

Where: ET _i — crop water requirement on day i;

ET _0i — reference crop water requirement on day i;

K _ci — crop coefficient on the i-th day;

K _wi —Soil moisture correction coefficient on the i-th day when non-full irrigation is implemented;

s - cluster centers of different weather types;

Get the crop online real-time irrigation forecast model with days as the time period:

M _i ＝W _i-1 +P _0i +W _Ti -ET _i +K _i -W _i

The online real-time irrigation forecast and control model controls the irrigation system to irrigate the crops in real time according to the real-time irrigation water volume M _i of the crops.

4. According to claim 3, the field intelligent irrigation online control management method is characterized in that, it also includes a daily recursive prediction model of the moisture content of the crop plan wetting layer, if the buried depth of groundwater in the research area is relatively deep, groundwater is ignored for crop replenishment Without timing, the water balance model of the crop plan wetting layer with days as the forecast period is:

W _i =W _i-1 +P _0i +W _Ti -ET _i +M _i ;

Among them, the soil water content of the planned wetting layer at the beginning and end of the i-th day is as follows:

W _i-1 ＝1000nH _i-1 θ _i-1

W _i ＝1000nH _i θ _i

Among them: H _i-1 — initial planned wet layer depth on the i-th day;

H _i —planned wetted layer depth at the end of day i;

θ _i-1 —Initial soil moisture content on the i-th day, expressed as a percentage of soil volume;

θ _i —soil moisture content at the end of the i-th day, expressed as a percentage of soil volume;

n—soil porosity;

The amount of water W _Ti increased due to the increase of the planned wetting layer in the i-th period;

W _Ti ＝1000(H _i -H _i-1 )·n·θ _deep

Where: θ _deep — deep soil moisture content;

The daily recursive prediction model of the moisture content of the crop plan wetting layer can be obtained:

{θ θ}_{i i} = = \frac{{H h}_{i i - - 11}}{{H h}_{i i}} {θ θ}_{i i - - 11} - - (({ET ET}_{i i - - 11} - - {P P}_{00 i i - - 11} - - {W W}_{Ti Ti} - - {M m}_{i i - - 11})) / / ((10001000 n no {H h}_{i i}));;

According to the day-by-day recursive prediction model of the moisture content of the wet layer in the crop plan, the soil moisture in the root zone of the crop is predicted day by day from the planting date of the crop until the crop is harvested, so as to carry out the day-by-day recursion of the soil moisture status during the entire growth period of the crop predict.

5. according to the described field intelligent irrigation online control management method of claim 3, it is characterized in that, also comprise the fuzzy clustering prediction method of crop water demand ET ₀ : carry out fuzzy clustering to historical long series ET ₀ under different weather types, obtain The cluster center s of different weather types performs self-correction of ET ₀ : ET′ ₀ =s·ET ₀ , to achieve more accurate prediction of crop water demand under future weather scenarios;

The cluster center s is determined by the method:

Let x _ij be the ET ₀ value under different weather types, wherein, i represents the weather type, i=1, 2, 3, 4, respectively represent sunny, cloud, overcast, rain; j represents ten days, j=1, 2, ..., 36; from the ET ₀ classification statistics table, the eigenvalue matrix of the multi-year average ET ₀ under different weather types in each ten-day of the year can be obtained:

{x x}_{ij ij} = = [\begin{matrix} {x x}_{1111} & {x x}_{1212} & . . . . . . & {x x}_{11 j j} \\ {x x}_{21 twenty one} & {x x}_{21 twenty one} & . . . . . . & {x x}_{22 j j} \\ . . . . . . & . . . . . . & . . . . . . & . . . . . . \\ {x x}_{i i 11} & {x x}_{i i 22} & . . . . . . & {x x}_{ij ij} \end{matrix}]

The relative membership degree matrix of ET ₀ eigenvalues in each ten-day period:

Among them: maxx _ij - the maximum value of the eigenvalue of ET ₀ in j days;

Then the relative membership degree matrix under the i-th weather type in j days is:

R R = = {r r}_{ij ij} = = [\begin{matrix} {r r}_{1111} & {r r}_{1212} & . . . . . . & {r r}_{11 j j} \\ {r r}_{21 twenty one} & {r r}_{22 twenty two} & . . . . . . & {r r}_{22 j j} \\ . . . . . . & . . . . . . & . . . . . . & . . . . . . \\ {r r}_{i i 11} & {r r}_{i i 22} & . . . . . . & {r r}_{ij ij} \end{matrix}]

Let the fuzzy cluster center matrix of ET ₀ under i weather types be:

S S = = (({s the s}_{ih i h})) = = [\begin{matrix} {s the s}_{1111} & {s the s}_{1212} \\ {s the s}_{21 twenty one} & {s the s}_{22 twenty two} \\ . . . . . . & . . . . . . \\ {s the s}_{i i 11} & {s the s}_{i i 22} \end{matrix}]

Among them: s _ih —— is the normalized number of eigenvalues of ET ₀ under i weather types under the classification mode h, 0≤s _ih ≤1;

h=1,2,...,c;

Assuming that the classification is carried out according to the relative membership degree of the sunny day ET ₀ , that is, c=2, then its fuzzy clustering matrix is:

U u = = (({u u}_{hj hj})) = = [\begin{matrix} {u u}_{1,1 1,1} & {u u}_{1,2 1,2} & . . . . . . & {u u}_{1,36 1,36} \\ {u u}_{2,1 2,1} & {u u}_{2,1 2,1} & . . . . . . & {u u}_{2,36 2,36} \end{matrix}]

And meet the conditions:

\begin{matrix} 00 \leq \leq {u u}_{hj hj} \leq \leq 11 \\ {Σ Σ}_{h h = = 11}^{c c} {u u}_{hj hj} = = 11 \\ {Σ Σ}_{j j = = 11}^{n no} {u u}_{hj hj} > > 00 \end{matrix}\}

Then the difference between the j-day sample and the ET ₀ cluster center h of different weather types can be expressed by the generalized Euclidean weighted distance, namely

| | | | {w w}_{i i} (({r r}_{i i} - - {s the s}_{h h})) | | | | = = {{{Σ Σ}_{i i = = 11}^{m m} [[{w w}_{i i} {(({r r}_{ij ij} - - {s the s}_{ih i h}))}^{22}]]}}^{\frac{11}{22}}

Where: i—the total number of weather types;

The weighted generalized Euclidean weighted distance between sample j and category h is obtained by taking u _hj as weight:

F(u _hj )＝u _hj ||w _i (r _j -s _h )||,

Calculate the cluster center s, and get the corrected value ET′ ₀ of ET ₀ .

6. according to claim 5 described field intelligent irrigation online control management method, it is characterized in that, optimal fuzzy clustering center matrix s solution step is specifically as follows:

(1) Given the calculation accuracy ε ₁ and ε ₂ required by u _hj and _sih ;

(2) Assume an initial fuzzy clustering matrix that satisfies the constraints u _hj and the elements are not all equal

(3) substitute The corresponding initial fuzzy clustering center matrix

{s the s}_{ih i h}^{00} = = \frac{{Σ Σ}_{j j = = 11}^{n no} {u u}_{hj hj}^{00}^{22} {w w}_{i i}^{22} {r r}_{ij ij}}{{Σ Σ}_{j j = = 11}^{n no} {u u}_{hj hj}^{00}^{22} {w w}_{i i}^{22}}

(4) substitute Find an approximate fuzzy clustering matrix

{u u}_{hj hj}^{11} = = \frac{11}{{Σ Σ}_{h h = = 11}^{c c} \frac{{Σ Σ}_{i i = = 11}^{n no} {[[{w w}_{i i} (({r r}_{ij ij} - - {s the s}_{ih i h}^{00}))]]}^{22}}{{Σ Σ}_{i i = = 11}^{m m} {[[{w w}_{i i} (({r r}_{ij ij} - - {s the s}_{ih i h}))]]}^{22}}}

(5) substitute Finding an approximate fuzzy clustering center matrix

(6) Compare the matrices one by one and and the matrix The corresponding elements of , if and then the iteration ends, is the optimal cluster center matrix s, otherwise Repeat the steps from (3) to (5) until the accuracy _requirements are met. Existing studies have proved the convergence of iterations; through fuzzy clustering calculations, the fuzziness of ET 0 in cloudy, overcast and rainy weather to ET ₀ in sunny days can be obtained cluster center.

7. According to claim 3, the field intelligent irrigation online control management method is characterized in that, sufficient or insufficient irrigation is carried out according to crop water conditions, and when the incoming water is large enough to meet the irrigation demand, the amount of irrigation water M _i is:

M _i ＝1000·n·H _i (1-θ _i )·θ _max

Among them: θ _i — the initial soil moisture content of the i-th day (stage);

H _i —the depth of the crop plan wet layer on the i-th day (stage);

When the inflow of water is insufficient or the amount of water resources is in short supply, the crops are under-irrigated; the amount of irrigation water M _i is:

M _i ＝1000·n·H _i (θ _c1 -θ _i )·θ _max

Among them: θ _c1 — the soil moisture content to be achieved after irrigation, and the specific value is determined according to the crop tolerance in the case of insufficient irrigation.

8. According to claim 3, the field intelligent irrigation online control management method is characterized in that, the layered measurement of soil moisture and the real-time correction of soil moisture;

This system regards the soil layer above the root activity area of crops as a whole system, and selects the measured soil water moisture content at four monitoring times of 2:00, 8:00, 14:00, and 20:00 from the monitored soil moisture data, and takes the crop The average value of the soil moisture content measured by all sensors above the depth of the planned wet layer at different times is taken as the actual soil moisture value of the day;

Let θw _i (j, h _l ) represent the soil water content measured by the sensor at the depth h _l of the soil layer at the monitoring time j on the i-th day, then at 2 o'clock, 8 o'clock, 14 o'clock, The soil moisture values at depth h _l can be recorded as: θw _i (2, h _l ), θw _i (8, h _l ), θw _i (14, h _l ), θw _i (20, h _l ) , then the soil moisture value at a given soil depth h _l on the i-th day can be calculated using the arithmetic mean method

{θ θ}_{i i,, {h h}_{l l}} = = \frac{θ θ {w w}_{i i} ((22,, {h h}_{l l})) + + θ θ {w w}_{i i} ((88,, {h h}_{l l})) + + θ θ {w w}_{i i} ((1414,, {h h}_{l l})) + + θ θ {w w}_{i i} ((2020,, {h h}_{l l}))}{44}

The soil moisture value θ _i of the planned wetting layer on the i-th day can be obtained by using the weighted mean method:

{θ θ}_{i i} = = {Σ Σ}_{l l = = 11}^{m m} {w w}_{i i} {θ θ}_{i i,, {h h}_{l l}}

In the formula: w _i — is the soil moisture value of the soil layer depth hl The influence weight on θ _i ;

m—the number of soil layers;

When predicting soil moisture content, days are used as the calculation period, and the recursive process is as follows: when the growth period is running for the first time, the soil moisture content is measured once at the beginning of the first day of the growth period, and it is used as the initial value of the stage. Recursion formula (6) recursively calculates the soil moisture content at the end of each day, compares and corrects it with the actual measured soil moisture content of the day, and takes the corrected soil moisture content as the initial value of the next stage, so the order of the day Recursively, carry out the simulation and correction of the moisture content of the wet layer in the daily plan;

In the process of calculation, if the moisture content on the i-th day is estimated to be less than or equal to the minimum allowable moisture content in the growth period of the crop, and taking into account the weather forecast, in the case of no rain or very little rainfall, use Section 5.2 The real-time irrigation forecast model makes irrigation forecast for this day and publishes it on the network. The soil moisture content of the day is corrected to the soil moisture content after irrigation, and then the moisture content of the planned wet layer is deduced from this as the initial value until the growth period End; if there is rainfall in the weather forecast, the effective rainfall needs to be considered, and the soil moisture is used to recursively day by day

θ_{i} = \frac{h_{i - 1}}{h_{i}} θ_{i - 1} - ({ET}_{i - 1} - P_{0 i - 1} - W_{Ti} - m_{i - 1}) / (1000 no h_{i})

Analyze and predict the moisture content of the planned wetting layer. If heavy rainfall occurs during the crop growth period, causing the soil moisture content to exceed the field water holding rate during the simulation calculation process, the soil moisture content will be treated as field water holding capacity rate, the excess water seeps into the deep soil.

9. according to the described field intelligent irrigation online control management method of claim 3, it is characterized in that, the real-time correction model of the crop coefficient is In the formula, ET′ _i-1 is the actual water demand of crops on day i-1 after correction, and the calculation formula is: ET′ _i-1 ＝W _i-1 +P _0i-1 +W′ _ri +M _i-1 -W′ _i , K′ _w,i-1 is the actual soil moisture correction coefficient on the i-1 day, W _i-1 and W _i-1 ' are the soil moisture content at the beginning and end of the i-1 day, respectively Water volume, M _i-1 is the irrigation water volume of the i-1 day, P _0i-1 is the effective rainfall of the i-1 day;

At the end of day i-1, the measured initial value of soil moisture θ _i on day _i is known; if the initial value of soil moisture θ _i ' predicted on day i is very close to the measured initial value of soil moisture The crop coefficient of -1 day takes the initial value; if the difference between θ _i ' and θ _i is relatively large, the actual water demand of the crop on the i-1 day can be deduced from the measured soil moisture value θ _i :

ET′ _i-1 ＝1000nH _i-1 θ _i-1 +P _0i-1 +W′ _ri +M _i-1 -1000nH _i θ _i

Among them: ET _i-1 ′—the actual water requirement of crops on day i-1 after correction;

M _i-1 — irrigation amount on day i-1;

Then the actual water requirement of crops on day i-1 after correction is:

ET′ _i-1 ＝W _i-1 +P _0i-1 +W′ _ri +M _i-1 -W′ _i

In the formula: W _i-1 , W′ _i —— soil water content at the beginning and end of day i-1, respectively;

Then the corrected i-1 day crop coefficient K′ _ci-1 can be obtained;

{K K}_{{c c,,}_{i i - - 11}}^{' '} = = {ET ET}_{i i - - 11}^{' '} / / (({K K}_{{w w,,}_{i i - - 11}}^{' '} \cdot &Center Dot; {ET ET}_{{00,,}_{i i - - 11}}))

That is, the crop coefficient value of the day is corrected at the end of the i-1 day, and the corrected value is used as the input value of the next calculation period, and so on, so as to realize the daily correction of the crop coefficient K _c value in the whole growth period.

10. The field intelligent irrigation online control management method according to claim 3, characterized in that,

Also included is a model for determining the initial value of the crop-planned wetting depth:

{H h}_{i i} = = {h h}_{n no - - 11} + + (({h h}_{n no} - - {h h}_{n no - - 11})) \cdot &Center Dot; ((i i - - {Σ Σ}_{j j = = 11}^{n no} {l l}_{j j - - 11}^{j j - - 11}_{j j - - 11})) / / {l l}_{n no}^{n no}

Among them: H _i ——planned wetting layer depth on the i-th day (stage) of the crop;

h _n-1 ——planned wetting layer depth at the beginning of the nth growth period;

h _n ——planned wetting layer depth at the end of the nth growth period;

n - the growth period of the crop;

i - cumulative days of crop growth after sowing;

- the number of days of growth in the nth growth period;

——the growth days of the jth growth period, j=1,2,...,n.