CN-122004114-A - Agricultural intelligent irrigation method and system based on big data analysis

CN122004114ACN 122004114 ACN122004114 ACN 122004114ACN-122004114-A

Abstract

The invention discloses an agricultural intelligent irrigation method and system based on big data analysis, and relates to the technical field of agricultural informatization and irrigation control, comprising the following steps of collecting irrigation related data, preprocessing the irrigation related data to obtain a standardized irrigation characteristic vector; the method comprises the steps of carrying out water demand prediction on standardized irrigation feature vectors by utilizing a BP neural network to obtain a water demand predicted value, carrying out collaborative optimization calculation on irrigation control parameters by adopting a group parameter collaborative adjustment method based on the water demand predicted value to obtain an optimal irrigation strategy instruction set, and carrying out control treatment on irrigation execution parameters according to the optimal irrigation strategy instruction set to obtain a corresponding irrigation execution control result. The invention realizes the automation and the intellectualization of the whole irrigation flow from data perception, intelligent prediction and global optimization to precise execution, and remarkably improves the scientificity of irrigation decision, the utilization efficiency of water resources and the self-adaptive capacity of the system.

Inventors

Zhan Lichuan
DENG CANWU
CUI HUIXIA
Duan Bojun
XU XIN

Assignees

洛阳德道农业科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260409

Claims (10)

1. An agricultural intelligent irrigation method based on big data analysis is characterized by comprising the following steps: step S1, collecting irrigation related data, and preprocessing the irrigation related data to obtain a standardized irrigation characteristic vector; s2, carrying out water demand prediction on the standardized irrigation characteristic vector by using a BP neural network to obtain a water demand predicted value; step S3, based on the water demand pre-estimated value, adopting a group parameter cooperative adjustment method to perform cooperative optimization calculation on irrigation control parameters to obtain an optimal irrigation strategy instruction set; And S4, controlling and processing the irrigation execution parameters according to the optimal irrigation strategy instruction set to obtain a corresponding irrigation execution control result.
2. An agricultural intelligent irrigation method based on big data analysis as claimed in claim 1, wherein said step S1 comprises: Step S101, collecting irrigation related data through a sensor network; The sensor network comprises a soil humidity sensor, a soil temperature sensor, an illumination intensity sensor, an air temperature sensor, an air humidity sensor, a wind speed sensor, a wind direction sensor, a rainfall sensor and a crop canopy temperature sensor; the irrigation related data comprise soil humidity, soil temperature, illumination intensity, air temperature, air humidity, wind speed, wind direction, rainfall and crop canopy temperature; step S102, data cleaning and abnormal value processing are carried out on irrigation related data to obtain cleaned data; Step S103, calculating soil humidity, soil temperature, illumination intensity, air temperature, air humidity, wind speed, wind direction, rainfall and crop canopy temperature in the cleaned data by using a crop water stress model to obtain crop water stress index values; step S104, normalizing the cleaned data and the crop water stress index value to obtain normalized data; and step S105, extracting water demand prediction features of the normalized data to obtain a standardized irrigation feature vector.
3. The agricultural intelligent irrigation method based on big data analysis as claimed in claim 2, wherein in step S105, the water demand prediction feature extraction is performed on the normalized data, specifically including: extracting current real-time values of soil humidity, soil temperature, illumination intensity, air temperature, air humidity, wind speed, wind direction, rainfall and crop canopy temperature in the normalized data; Calculating an average value of the normalized data in a time window based on a preset time window; and combining the current real-time value and the average value to obtain the standardized irrigation characteristic vector.
4. The agricultural intelligent irrigation method based on big data analysis according to claim 1, wherein the processing logic for performing water demand prediction on the standardized irrigation feature vector by using the BP neural network comprises: Normalized irrigation eigenvector Each component of (3) Endowing the neural network with corresponding nodes of the input layer; For the first The calculation formulas of the input values and the output values of the hidden layer nodes are as follows: ; ; Wherein, the Represent the first The input values of the nodes of the hidden layer, Representing the number of the input layer node, Representing the pre-training determination, from input layer From the node to the th The connection weights of the nodes of the hidden layer, Represent the first The individual hidden layer nodes pre-train the determined bias threshold, Represent the first The output values of the nodes of the hidden layer, Representing a predefined activation function that is to be activated, Representing the total number of input layer nodes; The calculation formula of the input value and the water demand predicted value of the output layer node is as follows: ; ; Wherein, the Representing the input value of the output layer node, Representing the pre-training determination, from the first The connection weights of the hidden layer nodes to the output layer nodes, Representing a pre-trained determined bias threshold for the output layer node, The water demand pre-estimated value is indicated, Representing the output layer activation function.
5. The agricultural intelligent irrigation method based on big data analysis according to claim 1, wherein the processing logic for performing collaborative optimization calculation on irrigation control parameters by adopting a group parameter collaborative adjustment method based on the water demand pre-estimated value comprises: based on the water demand predicted value, minimizing deviation between the actual irrigation total amount and the predicted water demand to be an optimization target, and constructing an objective function; Determining a search space of irrigation duration and flow parameters based on the water demand pre-estimated value; generating an initial irrigation strategy population in the search space, each irrigation strategy individual in the irrigation strategy population being composed of irrigation control parameters; The irrigation control parameters comprise irrigation starting time, irrigation duration and irrigation flow distribution proportion; based on a group parameter collaborative adjustment method, group state distribution information and historical strategy memory information are introduced, and the parameter states of irrigation strategy individuals are iteratively updated until preset termination conditions are met, and an optimal irrigation strategy instruction set is output; The population state distribution information comprises distribution information entropy reflecting the degree of distributed density of population individuals in a search space; the history policy memorization information comprises history policy memorization guide vectors obtained based on history elite policy position vector fusion; the parameter state is a position vector in a search space.
6. The agricultural intelligent irrigation method based on big data analysis according to claim 5, wherein the group parameter collaborative adjustment method comprises a diversity seeding strategy, a history strategy memory optimization operator and a drought responsive disturbance mechanism; the diversity seeding strategy is used for uniformly exploring and initializing the search space to obtain an initial irrigation strategy population, and the calculation formula is as follows: ; Wherein, the Representing an initial strategy population The entropy of the distribution information in the search space, Representing the total number of discrete sub-regions dividing the solution space, Indicating that the population individuals fall into the first Probability density estimation of the individual regions, Indicating the index number region.
7. The agricultural intelligent irrigation method based on big data analysis according to claim 6, wherein the historical policy memory optimization operator performs fractional order integral memory fusion calculation on the historical elite policy position vectors to obtain a historical policy memory guide vector, and the calculation formula is: ; Wherein, the Represent the first The generation history policy memorizes the boot vector, Representing the gamma function of the gamma ray, Representing the memory decay factor and satisfying , Representing history of the first The elite strategy position vector of the generation archive, Representing the current number of iterations.
8. The agricultural intelligent irrigation method based on big data analysis according to claim 6, wherein the drought responsive disturbance mechanism specifically comprises: the drought response factor is calculated based on the water demand predicted value and the preset crop water demand standard value, and the calculation formula is as follows: ; Wherein, the Represent the first Drought response factors at the time of the iteration, Represent the first The water demand pre-estimated value at the time of iteration, Representing a preset water demand standard value of crops; Judging whether the drought response factor meets a preset trigger condition ; Wherein, the Representing a preset drought trigger threshold; if the preset trigger condition is met, executing disturbance update operation; Disturbance updating operation is carried out on irrigation strategy individuals in the current irrigation strategy population; The disturbance updating operation determines disturbance amplitude according to drought response factors, performs weighting processing on random disturbance vectors by using the disturbance amplitude to obtain target disturbance vectors, and superimposes the target disturbance vectors on current position vectors of irrigation strategy individuals to obtain updated irrigation strategy individual position vectors, wherein the calculation formula is as follows: ; Wherein, the Representing the updated irrigation strategy individual position vector, Represent the first The first iteration The position vector of the individual irrigation strategies, Representing the disturbance amplitude control parameter, Representing random disturbance vectors obeying a preset distribution; If the preset trigger condition is not met, keeping the current position vector of the current irrigation strategy individual unchanged, and taking the position vector as an initial state of next iteration evolution.
9. The intelligent agricultural irrigation method based on big data analysis according to claim 1, wherein the control processing is performed on the irrigation execution parameters according to the optimal irrigation strategy instruction set to obtain corresponding irrigation execution control results, and the processing logic comprises: analyzing the irrigation starting time, the irrigation duration and the irrigation flow distribution proportion in the optimal irrigation strategy instruction set; Performing time scale conversion on the water demand predicted value in the irrigation duration, performing dynamic weight distribution in combination with irrigation flow distribution proportion, and calculating to obtain a real-time execution flow control sequence; taking irrigation starting time as a time axis reference origin, discretizing and mapping a real-time execution flow control sequence on a time axis, and establishing a corresponding relation between each time step and the flow output intensity of the equipment; triggering an irrigation task at the irrigation starting time, and adjusting irrigation equipment according to the real-time execution flow control sequence to obtain a corresponding irrigation execution control result; the irrigation execution control result comprises a corresponding flow output value, irrigation starting time, irrigation ending time and accumulated irrigation water quantity in the continuous irrigation process under each time step.
10. An agricultural intelligent irrigation system based on big data analysis is applied to the agricultural intelligent irrigation method based on big data analysis as set forth in any one of claims 1-9, and is characterized by comprising a data processing module, a water demand prediction module, a cooperative optimization module and an irrigation execution module; the data processing module is used for collecting irrigation related data, preprocessing the irrigation related data and obtaining standardized irrigation characteristic vectors; The water demand prediction module is used for inputting the standardized irrigation characteristic vector into the BP neural network to perform water demand prediction so as to obtain a water demand predicted value; The collaborative optimization module is used for carrying out collaborative optimization calculation on irrigation control parameters by adopting a group parameter collaborative adjustment method based on the water demand predicted value to obtain an optimal irrigation strategy instruction set; the irrigation execution module is used for controlling and processing irrigation execution parameters according to the optimal irrigation strategy instruction set to obtain corresponding irrigation execution control results.

Description

Agricultural intelligent irrigation method and system based on big data analysis Technical Field The invention relates to the technical field of agricultural informatization and irrigation control, in particular to an agricultural intelligent irrigation method and system based on big data analysis. Background In recent years, water resource shortage and inefficiency are global challenges that limit sustainable development of agriculture. Traditional irrigation modes depend on a fixed irrigation system, artificial experience or simple feedback control based on a single-point sensor, are difficult to adapt to the high heterogeneity of water demand of crops on a space-time scale and dynamic fluctuation of weather and soil conditions, often cause insufficient irrigation or excessive irrigation, and cause water resource waste, energy consumption increase and potential negative environmental influence. At present, in China patent of the invention with the publication number of CN111524024A, an intelligent water and fertilizer irrigation system and an analysis method based on big data are disclosed, the system comprises a water and fertilizer irrigation data center, a water and fertilizer irrigation model uploading center, a water and fertilizer irrigation operation model analysis center, an intelligent water and fertilizer integrated irrigation module and a pipeline disinfection module, on the basis that the traditional water and fertilizer irrigation system solves the problem of fertilization and watering, further analysis and discussion are further conducted, the big data analysis technology, the AI technology and the Internet of things technology are combined, upgrading and optimizing are conducted on the water and fertilizer irrigation operation mode of a traditional depending technician, and the water and fertilizer irrigation data model is collected, analyzed, processed and applied through the big data processing technology, so that intelligent agriculture is realized. However, the related technology lacks collaborative optimization and dynamic adjustment of irrigation control parameters, lacks a mechanism for intelligent optimization by using historical data, and does not have an effective mechanism for coping with extreme weather (such as drought). Disclosure of Invention The invention solves the technical problems that the dynamic collaborative optimization of irrigation control parameters is difficult to realize in the prior art, and particularly the real-time adjustment of irrigation starting time, duration and flow distribution proportion cannot fully cope with the space-time heterogeneity of crop water demand, meteorological variation and soil conditions. The prior art also lacks a mechanism for effectively utilizing historical irrigation data to carry out strategy optimization, and when the system faces extreme climates such as drought, the irrigation strategy cannot be automatically adjusted, so that water resource waste or insufficient water supply of crops are caused. In order to solve the technical problems, the invention provides the following technical scheme that in the first aspect, the agricultural intelligent irrigation method based on big data analysis comprises the following steps: step S1, collecting irrigation related data, and preprocessing the irrigation related data to obtain a standardized irrigation characteristic vector; s2, carrying out water demand prediction on the standardized irrigation characteristic vector by using a BP neural network to obtain a water demand predicted value; step S3, based on the water demand pre-estimated value, adopting a group parameter cooperative adjustment method to perform cooperative optimization calculation on irrigation control parameters to obtain an optimal irrigation strategy instruction set; And S4, controlling and processing the irrigation execution parameters according to the optimal irrigation strategy instruction set to obtain a corresponding irrigation execution control result. As a preferable scheme of the agricultural intelligent irrigation method based on big data analysis, the step S1 includes: Step S101, collecting irrigation related data through a sensor network; The sensor network comprises a soil humidity sensor, a soil temperature sensor, an illumination intensity sensor, an air temperature sensor, an air humidity sensor, a wind speed sensor, a wind direction sensor, a rainfall sensor and a crop canopy temperature sensor; the irrigation related data comprise soil humidity, soil temperature, illumination intensity, air temperature, air humidity, wind speed, wind direction, rainfall and crop canopy temperature; step S102, data cleaning and abnormal value processing are carried out on irrigation related data to obtain cleaned data; Step S103, calculating soil humidity, soil temperature, illumination intensity, air temperature, air humidity, wind speed, wind direction, rainfall and crop canopy temperature in the cleaned data by usi