Intelligent irrigation measurement and control method and system for rice field fertilization and irrigation based on multi-source information fusion

Abstract

The application is suitable for the field of intelligent agricultural irrigation technology, and provides a rice field fertilizer and water intelligent irrigation measurement and control method and system based on multi-source information fusion, which comprises the following steps: according to the growth stage of rice seedlings, determining the target fertilizer and water demand amount through a quantitative relationship model of fertilizer and water input and crop growth response; based on an improved D-S evidence theory algorithm, fusing real-time multi-source monitoring data, dynamically evaluating the environmental state of the current planting area according to the target fertilizer and water demand amount, and combining future weather forecast data to generate irrigation decision results that take into account crop demand and resource conservation; based on a fuzzy PID control algorithm, automatically controlling fertilizer and water irrigation operation according to the irrigation decision results. The application collects and compares databases of the needs of crops themselves and growth periods for environmental water and nutrients, comprehensively decides the irrigation process according to various factors such as crop growth periods and weather, effectively improves the utilization efficiency of water resources and fertilizers, and has good application effect.

Description

Technical Field

[0001] This invention belongs to the field of intelligent agricultural irrigation technology, and in particular relates to a method and system for intelligent irrigation measurement and control of paddy fields based on multi-source information fusion. Background Technology

[0002] my country's water shortage problem is becoming increasingly severe, especially in Northeast China, where the contradiction between water supply and demand is prominent, becoming a key factor restricting the region's sustainable economic and social development. Meanwhile, agriculture accounts for as much as 75% of water use in the three northeastern provinces, but the traditional flood irrigation method achieves a water utilization rate of less than 50%, leading to serious waste. Therefore, vigorously developing intelligent and efficient water-saving irrigation technologies for agriculture and improving agricultural water use efficiency has become an inevitable choice to alleviate the water supply and demand contradiction in Northeast my country, and also has important practical significance for ensuring national water security and promoting sustainable agricultural development.

[0003] Water-saving irrigation refers to an irrigation model that optimizes water resource allocation and management technologies to achieve optimal agricultural production and economic benefits with minimal irrigation water consumption, while ensuring normal crop growth. The core principle of this irrigation concept lies in using systematic technological integration methods combined with scientific management strategies to significantly improve agricultural water use efficiency, ultimately achieving the goal of sustainable water resource utilization.

[0004] However, the methods and overall monitoring and control systems currently used in irrigation in China are relatively simple, unable to accurately control the water requirements of crops at different growth stages, and thus failing to meet environmental protection requirements. Therefore, it is necessary to improve the existing irrigation methods and monitoring and control systems. Summary of the Invention

[0005] The purpose of this invention is to provide a method for intelligent irrigation and control of paddy fields based on multi-source information fusion, in order to solve the above-mentioned technical problems.

[0006] This invention is implemented as follows: a smart irrigation monitoring and control method for paddy fields based on multi-source information fusion, comprising the following steps:

[0007] Based on the multi-dimensional information collected in the early stage, a quantitative relationship model between fertilizer and water input and crop growth response was constructed using deep belief networks.

[0008] Periodically collect real-time multi-source monitoring data of the planting area and determine the growth stage of rice seedlings;

[0009] Based on the growth stages of rice seedlings, the target fertilizer and water requirements are determined through a quantitative relationship model between fertilizer and water input and crop growth response.

[0010] Based on the improved DS evidence theory algorithm, real-time multi-source monitoring data is integrated to dynamically assess the environmental status of the current planting area according to the target fertilizer and water requirements, and combined with future weather forecast data to generate irrigation decision results that take into account both crop needs and resource conservation.

[0011] Based on the fuzzy PID control algorithm, the fertilization and irrigation operations are automatically controlled according to the irrigation decision results.

[0012] Preferably, the multi-dimensional information includes soil moisture data, meteorological environmental parameters, and crop growth status; the quantitative relationship model between fertilizer and water input and crop growth response is expressed as follows:

[0013] ;

[0014] Where Y is the output crop response variable; X i is the input multi-dimensional information vector; W is the weight matrix, optimized through backpropagation algorithm; b is the bias term.

[0015] Preferably, the multi-source monitoring data is collected periodically through soil parameter monitoring units, meteorological environment monitoring units, and crop growth monitoring units, based on a sensor network deployed in the planting area.

[0016] Preferably, the steps of generating irrigation decision results that balance crop needs and resource conservation, based on the improved DS evidence theory algorithm, integrating real-time multi-source monitoring data, dynamically assessing the current state of the planting area according to the target fertilizer and water requirements, and combining future weather forecast data, specifically include:

[0017] Define a framework for identifying the status of planting areas;

[0018] The basic weights are dynamically assigned based on the preset sensor type and environmental conditions, and the weight coefficients are dynamically adjusted based on future weather forecast data to obtain the dynamic weights of each type of data in the multi-source monitoring data.

[0019] Calculate the basic probability allocation function for each type of data based on the dynamic weights of each type of data;

[0020] Based on the basic probability allocation function of each type of data, real-time multi-source monitoring data are integrated to determine the state confidence level;

[0021] When conflicts occur among different types of data in multi-source monitoring data, the entropy weight method is used to correct the dynamic weights.

[0022] Based on the state confidence level, determine the water demand index and It then uses preset thresholds to make judgments and generate irrigation decision results.

[0023] Preferably, the rules for generating irrigation decision results are as follows:

[0024] Irrigation should be initiated when the water demand index = bel (water shortage) + 0.5 * Pl (water shortage) > 0.7;

[0025] When the fertilizer requirement index equals bel (nitrogen deficiency) When Pl (phosphorus deficiency) > 0.6, Where bel is the trust function and Pl is the likelihood function.

[0026] Preferably, the fuzzy PID control algorithm includes a dual closed-loop control structure: its outer loop is the target flow rate, which is calculated from the target fertilizer and water demand; the inner loop is for real-time adjustment of the solenoid valve opening and the fertilizer pump speed; the variable input expression of the fuzzy PID control algorithm is:

[0027] ;

[0028] Where t is time; e(t) is the flow deviation; ec(t) is the rate of change of deviation; Q set For target traffic; Q actual The actual flow rate is used; the PID parameters are dynamically adjusted based on the fuzzy results of the flow rate deviation and the rate of change of deviation.

[0029] Preferably, the mapping relationship output by the fuzzy PID control algorithm is as follows:

[0030] ;

[0031] Among them, K v N represents the opening degree of the solenoid valve. p U(t) represents the rotational speed of the fertilizer pump; u(t) represents the final control variable; the final control variable is calculated using the center of gravity method.

[0032] ;

[0033] Where, k p k i k d These are the PID parameters, representing the proportional coefficient, integral coefficient, and derivative coefficient, respectively.

[0034] Preferably, the intelligent irrigation monitoring and control method for paddy fields based on multi-source information fusion further includes the following steps:

[0035] Remotely monitor fertigation operations and optimize the model based on actual irrigation execution parameters and effect data.

[0036] Another objective of this invention is to provide a smart irrigation monitoring and control system for paddy fields based on multi-source information fusion, used in the aforementioned smart irrigation monitoring and control method for paddy fields. The system includes a multi-source information acquisition system, a computer system, and a mechanical execution system. The multi-source information acquisition system is used to periodically collect real-time multi-source monitoring data of the planting area and determine the growth stage of the rice seedlings. The computer system includes an edge computing gateway and an intelligent control system.

[0037] The edge computing gateway includes:

[0038] The model building module is used to construct a quantitative relationship model between fertilizer and water input and crop growth response based on multi-dimensional information collected in the early stage and deep belief network.

[0039] The target input determination module is used to determine the target fertilizer and water requirements based on the growth stage of rice seedlings through a quantitative relationship model between fertilizer and water input and crop growth response.

[0040] Data fusion and irrigation decision module: Based on the improved DS evidence theory algorithm, it integrates real-time multi-source monitoring data, dynamically assesses the environmental status of the current planting area according to the target fertilizer and water requirements, and combines future weather forecast data to generate irrigation decision results that take into account both crop needs and resource conservation.

[0041] The intelligent control system includes:

[0042] The automatic control module is used to automatically control the fertigation operation based on the irrigation decision results using a fuzzy PID control algorithm.

[0043] Preferably, the intelligent irrigation and control system for paddy fields based on multi-source information fusion further includes:

[0044] The cloud-based remote monitoring and optimization system is used to remotely monitor fertigation operations and optimize the model based on actual irrigation execution parameters and effect data.

[0045] The present invention provides a method for intelligent irrigation and control of paddy fields based on multi-source information fusion. It establishes a comparative database by collecting data on the crop's own characteristics and its growth cycle requirements for environmental water and nutrients, and makes decisions on the irrigation process based on multiple factors such as the crop growth cycle and weather. By accurately collecting meteorological information parameters, soil moisture parameters, and water quantity parameters around the crop during its growth cycle, it intelligently compares the irrigation amount and fertilizer application amount to meet the suitable growth environment for rice, thereby achieving precise regulation of fertilizer and water in rice production in Northeast China, reducing costs and increasing yield. Attached Figure Description

[0046] Figure 1 This is a schematic diagram of the structure of the intelligent irrigation and control system for paddy fields based on multi-source information fusion, provided in an embodiment of the present invention.

[0047] Figure 2 This is a flowchart illustrating the intelligent irrigation and control method for paddy fields based on multi-source information fusion, as provided in an embodiment of the present invention.

[0048] Figure 3 This is a block diagram of a crop growth stage detection model provided in an embodiment of the present invention.

[0049] Figure 4 This is a flowchart illustrating step S400 in the intelligent irrigation and control method for paddy fields based on multi-source information fusion provided in an embodiment of the present invention.

[0050] Figure 5 This is a structural block diagram of a fuzzy PID controller provided in an embodiment of the present invention.

[0051] In the diagram: 1. Multi-source information acquisition system; 1-1. Soil parameter monitoring unit; 1-1-1. Water depth sensor; 1-1-2. Moisture content sensor; 1-1-3. Nutrient content sensor; 1-2. Meteorological environment monitoring unit; 1-2-1. Temperature sensor; 1-2-2. Humidity sensor; 1-2-3. Wind speed sensor; 1-2-4. Heat flux sensor; 1-2-5. Net radiation sensor; 1-3. Crop growth monitoring unit; 1-3-1. Multispectral imager; 1-4. Wireless data transmission module; 2. Computer system; 2-1. Edge computing gateway; 2-2. Intelligent control system; 3. Mechanical execution system; 3-1. Water pump; 3-2. Irrigation valve; 3-3. Fertilizer valve; 3-4. Alarm. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0053] This invention proposes a method and system for intelligent irrigation monitoring and control of paddy fields based on multi-source information fusion. By establishing a quantitative relationship model between fertilizer and water input and crop growth response, and combining real-time collected multi-source data such as soil moisture and water and fertilizer status, an improved DS evidence theory algorithm is used for intelligent decision-making to achieve precise irrigation control of paddy fields. The system innovatively introduces an adaptive weight adjustment mechanism to effectively improve the fusion accuracy of multi-source heterogeneous data, and achieves remote monitoring and dynamic optimization of irrigation strategies through a cloud platform, forming a complete intelligent irrigation closed-loop control system.

[0054] In practical applications, the above-mentioned intelligent irrigation and control method and system for paddy fields can be applied to the rice planting and production process, but it is not limited to this and can also be applied to the planting of other crops.

[0055] Specifically, such as Figure 1 As shown, taking the rice planting and production process in Northeast China as an example, in one embodiment of the present invention, a smart irrigation and control system for paddy fields based on multi-source information fusion is provided, which includes a multi-source information acquisition system 1, a computer system 2, and a mechanical execution system 3; the multi-source information acquisition system 1 is connected to the computer system 2 and the mechanical execution system 3 in sequence; the mechanical execution system 3 includes, but is not limited to, a water pump 3-1, an irrigation valve 3-2, a fertilizer valve 3-3, and an alarm 3-4, etc.

[0056] The multi-source information acquisition system 1 is used to periodically collect real-time multi-source monitoring data of the planting area and determine the growth stage of rice seedlings. Specifically, it includes a soil parameter monitoring unit 1-1, a meteorological environment monitoring unit 1-2, a crop growth monitoring unit 1-3, and a data wireless transmission module 1-4. In practical applications, the soil parameter monitoring unit 1-1 is a farmland sensor matrix, which may include, but is not limited to, a water depth sensor 1-1-1, a moisture content sensor 1-1-, and a nutrient content sensor 1-1-3. The meteorological environment monitoring unit 1-2 is a small weather station, which may include, but is not limited to, a temperature sensor 1-2-1, a humidity sensor 1-2-2, a wind speed sensor 1-2-3, a heat flux sensor 1-2-4, and a net radiation sensor 1-2-5. The crop growth monitoring unit 1-3 is an agricultural plant protection drone, which may include, but is not limited to, a multispectral imager 1-3-1. The data wireless transmission module 1-4 can use 4G / 5G wireless data transmission to realize the real-time transmission of multi-source farmland information.

[0057] The computer system 2 includes an edge computing gateway 2-1 and an intelligent control system 2-2; the edge computing gateway 2-1 can integrate multi-source information and make decisions, specifically including:

[0058] The model building module is used to construct a quantitative relationship model between fertilizer and water input and crop growth response based on multi-dimensional information collected in the early stage and deep belief network.

[0059] The target input determination module is used to determine the target fertilizer and water requirements based on the growth stage of rice seedlings through a quantitative relationship model between fertilizer and water input and crop growth response.

[0060] The data fusion and irrigation decision module, based on an improved DS evidence theory algorithm, integrates real-time multi-source monitoring data, dynamically assesses the environmental status of the current planting area according to the target fertilizer and water requirements, and combines future weather forecast data to generate irrigation decision results that take into account both crop needs and resource conservation.

[0061] The intelligent control system 2-2 includes an automatic control module, which automatically controls the fertilization and irrigation operations based on the irrigation decision results using a fuzzy PID control algorithm. The intelligent control system 2-2 receives the optimal irrigation amount decision results and regulates the start and stop operations of the water pump 3-1, irrigation valve 3-2, fertilization valve 3-3, and alarm 3-4, thereby realizing intelligent irrigation of paddy fields based on multi-source information fusion.

[0062] In a preferred embodiment of the present invention, the intelligent irrigation monitoring and control system for paddy fields based on multi-source information fusion further includes: a cloud-based remote monitoring and optimization system for remotely monitoring irrigation operations and optimizing the model based on actual irrigation execution parameters and effect data. This remote monitoring and optimization system supports historical data storage and analysis, and can continuously optimize decision model parameters based on actual irrigation effects. Furthermore, through a mobile terminal application, users can view field environmental data and equipment operating status at any time, and receive early warning information in abnormal situations, thus achieving intelligent management of paddy field irrigation. Specifically, the cloud-based remote monitoring and optimization system includes:

[0063] Data communication module: Adopts 4G / 5G multi-mode communication technology to ensure stable connection between field equipment and cloud platform; supports MQTT and CoAP lightweight protocols to achieve low-power data transmission; built-in breakpoint resume mechanism to automatically cache local data when network is abnormal.

[0064] Cloud Data Center: Deploys a time-series database (InfluxDB) to store historical sensor data; establishes a relational database (MySQL) to manage device information and user permissions; adopts a distributed storage architecture with configurable data retention period (default 3 years).

[0065] Intelligent Analysis Engine: Integrates an online machine learning framework, supports incremental model updates; has a built-in anomaly detection algorithm (based on the 3σ principle) to identify equipment faults in real time; and generates multi-dimensional irrigation performance analysis reports for daily, weekly, and monthly periods.

[0066] Visual interactive platform: Provides a web-based management backend that supports displaying device distribution on an electronic map; develops a mobile APP (Android / iOS) with main functions including: real-time viewing of soil moisture heat maps, receiving irrigation alarm push notifications (soil moisture content exceeding threshold, etc.), and remote manual control of irrigation valves.

[0067] Security protection system: Data transmission uses AES-256 encryption; device access implements two-way certificate authentication; operation audit logs are established to record all critical operations.

[0068] Model optimization closed loop: Automatically compares expected irrigation effects with actual crop responses; collects field feedback data through edge computing nodes; automatically generates model optimization suggestions monthly, which are then deployed and updated after confirmation by agronomists.

[0069] like Figure 2 As shown, in another embodiment of the present invention, a method for intelligent irrigation monitoring and control of paddy fields based on multi-source information fusion is provided, which is implemented using the above-mentioned system and specifically includes the following steps:

[0070] S100. Based on the multi-dimensional information collected in the early stage, a quantitative relationship model between fertilizer and water input and crop growth response is constructed based on deep belief network (DBN).

[0071] S200: Periodically collect real-time multi-source monitoring data of the planting area and determine the growth stage of rice seedlings;

[0072] S300. Based on the rice seedling growth stage, determine the target fertilizer and water requirements through a quantitative relationship model between fertilizer and water input and crop growth response.

[0073] S400, based on the improved DS evidence theory algorithm, integrates real-time multi-source monitoring data, dynamically assesses the environmental status of the current planting area according to the target fertilizer and water requirements, and combines future weather forecast data to generate irrigation decision results that take into account both crop needs and resource conservation.

[0074] The S500 is based on a fuzzy PID control algorithm and automatically controls the fertilization and irrigation operations according to the irrigation decision results.

[0075] In a preferred embodiment of the present invention, the multi-dimensional information includes soil moisture data, meteorological environmental parameters, and crop growth status. In practical applications, soil moisture data includes, but is not limited to, indicators such as stratified soil moisture content and nutrient content; meteorological environmental parameters include, but are not limited to, elements such as temperature, humidity, and light intensity; crop growth status monitoring data includes, but is not limited to, growth indicators such as leaf surface temperature and canopy structure, as well as plant physiological status information obtained through multispectral imaging.

[0076] The quantitative relationship model between fertilizer and water input and crop growth response is expressed as follows:

[0077] ;

[0078] Where Y is the output crop response variable; X i The input is a multi-dimensional information vector (e.g., X1—soil moisture content, X2—nitrogen content, ..., X...). n—Accumulated temperature, etc.); W is the weight matrix, optimized through backpropagation algorithm; b is the bias term. For example, in practical applications, the water requirement thresholds for different irrigation methods at various key growth stages of rice paddies in Northeast China are shown in Table 1. The weights of the corresponding data sources can be adjusted according to the water requirements of the rice tillering and booting stages in the table. Of course, the fertilizer nutrient requirements of rice paddies at various key growth stages in Northeast China are also different, and the fertilizer requirement thresholds can be determined according to the actual situation, which will not be elaborated here.

[0079] Table 1. Water requirements of rice at different growth stages in Northeast China

[0080]

[0081] In a preferred embodiment of the present invention, the multi-source monitoring data is periodically collected through a soil parameter monitoring unit 1-1, a meteorological environment monitoring unit 1-2, and a crop growth monitoring unit 1-3, based on a sensor network deployed in the planting area. For example, it can be collected at 5-minute intervals via a LoRa wireless sensor network.

[0082] Specifically, the soil parameter monitoring unit 1-1 is used to monitor sensor data such as soil moisture, pH value, nitrogen, phosphorus and potassium content, as well as to acquire soil moisture content, nutrient content data at different depths, and soil profile moisture content (double-layer measurement at 0-20cm and 20-40cm).

[0083] Meteorological environment monitoring units 1-2 are used to collect meteorological parameters such as temperature, humidity, wind speed, precipitation, light intensity, and photosynthetically active radiation (PAR), as well as future weather forecast data (6-hour forecasts of temperature, humidity, and rainfall probability).

[0084] Crop growth monitoring units 1-3 are used to acquire growth indicators such as leaf surface temperature and canopy structure, as well as crop growth data through multispectral imaging; all data are transmitted to computer system 2 after preprocessing. Among them, the NDVI index and crop canopy temperature (infrared thermometry accuracy ±0.5℃) are acquired using a multispectral imager (wavelength range 450-900nm).

[0085] like Figure 3 As shown, in a preferred embodiment of the present invention, a preset field crop growth stage detection model based on machine vision can be used to determine the growth stage of rice seedlings. Specifically, the following steps are included: acquiring crop images of the paddy field through a multispectral imager 1-3-1, then obtaining the growth status of each rice seedling in the image one by one, judging the growth stage based on the crop growth stage detection model, and determining the growth stage with the most statistically obtained growth stage as the current growth stage of the crop.

[0086] like Figure 4As shown, in a preferred embodiment of the present invention, step S400 specifically includes the following steps: based on the improved DS evidence theory algorithm, integrating real-time multi-source monitoring data, dynamically assessing the current state of the planting area according to the target fertilizer and water requirements, and combining future weather forecast data to generate an irrigation decision result that takes into account both crop needs and resource conservation.

[0087] S410. Define a framework for identifying the state of the planting area; for example, in practical applications, define a framework for identifying the moisture content state. as follows:

[0088] ;

[0089] It should be noted that a similar identification framework can be used to define the nutrient (fertilizer) content status, such as nitrogen deficiency and phosphorus deficiency, which will not be elaborated here.

[0090] S420. Based on the preset sensor type and environmental conditions, the basic weights are dynamically allocated, and the weight coefficients are dynamically adjusted according to future meteorological forecast data to obtain the dynamic weights of each type of data in the multi-source monitoring data. For example, the dynamic weight configuration of some sensors is shown in Table 2. The weight coefficients can be calibrated through field test data (such as the test data of Wuchang City from 2020 to 2023).

[0091] Table 2 Sensor Dynamic Weight Configuration

[0092]

[0093] S430. Calculate the basic probability allocation function for each type of data based on the dynamic weights of each type of data.

[0094] S440. Based on the basic probability allocation function of each type of data, fuse real-time multi-source monitoring data to determine the state confidence level; for example, in practical applications, the state confidence level corresponding to the sensor data in soil parameter monitoring unit 1-1. The calculation formula is as follows:

[0095] ;

[0096] Among them, A j To identify the state within the framework; x represents real-time multi-source monitoring data, specifically sensor data from soil parameter monitoring unit 1-1 in this embodiment; w i Dynamic weights; is the ideal parameter for the j-th growth stage of the crop; k is the growth stage index in this step.

[0097] S450. When conflicts occur among different types of data in multi-source monitoring data (such as soil moisture sensor malfunctions during rainfall), the entropy weight method is used to correct the dynamic weights, as follows:

[0098] ;

[0099] In this step, α k E represents the adjusted weight for the k-th piece of evidence, where n is the total number of pieces of evidence. k Let m be the information entropy of the sensor set. k (A i ) corresponds to state A i The state confidence.

[0100] S460. Based on the state confidence level, determine the water demand index and It then uses preset thresholds to make judgments and generate irrigation decision results.

[0101] Specifically, the rules for generating irrigation decision results are as follows:

[0102] Irrigation should be initiated when the water demand index = bel (water shortage) + 0.5 * Pl (water shortage) > 0.7;

[0103] When the fertilizer requirement index equals bel (nitrogen deficiency) When Pl (phosphorus deficiency) > 0.6, ;

[0104] Where bel is the trust function, Pl is the likelihood function; 0.7 and 0.6 are empirically determined thresholds.

[0105] In this embodiment of the invention, the improved DS evidence theory algorithm can make real-time decisions on the water and fertilizer requirements of rice based on a decision knowledge base of multi-dimensional parameters such as real-time soil moisture, meteorological conditions and crop growth. In the decision-making process, it takes into account the water evaporation compensation needs under extreme weather conditions, and can achieve precise irrigation volume control and nutrient supply.

[0106] In a preferred embodiment of the present invention, the fuzzy PID control algorithm is implemented based on a fuzzy PID controller, the block diagram of which is shown below. Figure 5 As shown; specifically, the fuzzy PID control algorithm includes a dual closed-loop control structure: its outer loop is the target flow rate, which is calculated from the target fertilizer and water demand; the inner loop is for real-time adjustment of the solenoid valve opening and the fertilizer pump speed; the variable input expression of the fuzzy PID control algorithm is:

[0107] ;

[0108] Where t is time; e(t) is the flow deviation; ec(t) is the rate of change of deviation; Q set For target traffic; Q actual The actual flow rate is used; the PID parameters are dynamically adjusted based on the fuzzy results of the flow rate deviation and the rate of change of deviation, as shown in Table 3.

[0109] Table 3. Dynamically adjusting PID parameters based on fuzzification results.

[0110]

[0111] Among them, the fuzzy results of the flow deviation e(t) include S (very small), M (medium) and B (very large); the fuzzy results of the deviation change rate ec(t) include NB (large negative, rapid deterioration), ZO (zero, stable) and PB (large positive, rapid improvement).

[0112] In a preferred embodiment of the present invention, the mapping relationship output by the fuzzy PID control algorithm is as follows:

[0113] ;

[0114] Among them, K v N represents the opening degree of the solenoid valve. p The speed of the fertilizer pump is given; u(t) is the final control variable; the final control variable u(t) is calculated using the center of gravity method:

[0115] ;

[0116] Where, k p k i k d These are the PID parameters, representing the proportional coefficient, integral coefficient, and derivative coefficient, respectively.

[0117] Furthermore, in step S500, based on the optimal irrigation decision results generated above, the opening degrees of the solenoid valve and fertilizer pump can be automatically adjusted through the aforementioned fuzzy PID control algorithm, enabling the mechanical execution system 3 to perform precise irrigation operations. It is worth noting that during irrigation, by real-time monitoring of parameters such as actual irrigation volume and soil moisture changes, and by feeding the execution effect back to the intelligent control system 2-2, a closed-loop adjustment mechanism is formed to ensure that the irrigation accuracy meets the expected requirements.

[0118] In a preferred embodiment of the present invention, the above-mentioned intelligent irrigation monitoring and control method for paddy fields based on multi-source information fusion further includes the following steps: remotely monitoring irrigation operations and optimizing the model based on actual irrigation execution parameters and effect data. It is worth noting that during actual irrigation operations, the execution parameters and effect data for each irrigation can also be recorded simultaneously, providing a basis for the optimization of the above model.

[0119] The above-mentioned intelligent irrigation and control method for paddy fields based on multi-source information fusion is applied to the actual planting and production process of rice in Northeast China. In step S100, a quantitative relationship model between fertilizer and water input and crop growth response is constructed through a deep belief network (DBN). The dataset used for model training comes from a paddy field in Changchun City, Jilin Province (10 mu in area, black soil type, and "Daohuaxiang No. 2" planted). After training, the model can predict the output (i.e., the target fertilizer and water demand, including the target fertilizer application rate and the target irrigation rate) based on the input (i.e., real-time collected soil moisture data, meteorological environmental parameters, and crop growth status). The computer system 2 can be a Windows 10 system, and its specific environment configuration for model training is as follows. As shown.

[0120] Table 4 Computer system environment configuration for model training

[0121]

[0122] In step S200, a sensor array can be deployed in the field to collect multi-source monitoring data, specifically, data can be collected at a frequency of once per hour. For example, the soil sensor may collect data showing that the current soil moisture content (at a depth of 0-30cm) is 28% and the soil bulk density is 1.3g / cm³. 3 The fertility sensor collected data showing that the current nitrogen content in the field is 75 mg / kg; the crop sensor collected data through image recognition showing that the crop height is 35 cm, and combined with the rice growth pattern, it was determined that the rice crop is currently in the tillering stage; the meteorological sensor collected data such as no rainfall and a temperature of 18℃ on that day.

[0123] In steps S300 and S400, relevant data collected by different types of sensors are uploaded to the edge computing gateway 2-1 via 4G / 5G communication technology. Based on the quantitative relationship model between fertilizer and water input and crop growth response, the target fertilizer and water requirements for high-quality growth of rice in the tillering stage in the paddy field are calculated as follows: soil moisture content 30%~35%, nitrogen content 80~100 mg / kg, phosphorus content 80~100 mg / kg, with an error σ=1.5%. Then, using an improved DS evidence theory, after eliminating interference from factors such as soil temperature fluctuations, data fusion is performed, and it is determined that the current data is below the target value. The current state is defined as "water / fertilizer shortage," thus triggering regulatory calculations.

[0124] The irrigation amount is calculated as follows:

[0125] 28% × 10 mu × 667 m² / mu × 0.3 m (cultivated layer) × 1.3 g / cm³ × 1.0 (correction factor) ≈ 100 m 3 ;

[0126] The amount of fertilizer applied is calculated as follows:

[0127] 80mg / kg × 10 mu × 667m² / mu × 0.3m (cultivated layer) × 1000kg / m 3 ×0.12 (correction factor) ≈ 12 kg (nitrogen fertilizer);

[0128] 80mg / kg × 10 mu × 667m² / mu × 0.3m (cultivated layer) × 1000kg / m 3 ×0.12 (correction factor) ≈ 12 kg (phosphate fertilizer);

[0129] In step S500, the intelligent control system 2-2 based on the PID fuzzy control algorithm receives the above irrigation decision result (irrigation amount 100m³). 3 After receiving the instruction to apply 12kg of nitrogen fertilizer and 12kg of phosphorus fertilizer, the intelligent control system 2-2 first controls the water and fertilizer mixing device to mix the 12kg of nitrogen fertilizer and 12kg of phosphorus fertilizer with 100m³ of water according to the instruction. 3 The water and fertilizer are mixed, and after the mixture is complete, the electric valve will automatically open the irrigation pipeline to begin irrigation. A flow sensor monitors the irrigation volume in real time; when the actual irrigation volume reaches 100m³, the flow rate will be recorded. 3 Then, the signal is fed back to the intelligent control system 2-2, which then issues a closing command to control the electric valve to close automatically, thus completing the irrigation and fertilization operation.

[0130] In addition, the cloud-based remote monitoring and optimization system can activate the alarm 3-4 through the intelligent control system 2-2 when the computer system 2 loses electrical connection with a certain set of devices or when a certain detection parameter is abnormal during long-term use, thereby reminding the staff. The data wireless transmission module 1-4 connects to the cloud-based smartphone or computer to provide a second reminder.

[0131] In summary, as can be seen from the above implementation process, the intelligent irrigation and fertilization control method and system for paddy fields based on multi-source information fusion provided by the embodiments of the present invention can accurately regulate irrigation and fertilization according to the actual conditions of the paddy field, effectively improve the utilization efficiency of water resources and fertilizers, reduce labor costs, and has good application effects.

[0132] It should be noted that each of the above modules can be implemented as a computer program, which can run on a computer device. The computer device's memory can store the computer program that makes up each module, enabling the processor to execute each step of the above method.

[0133] It should be understood that although the steps in the flowcharts of the various embodiments of the present invention are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the various embodiments may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.

[0134] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Furthermore, any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory.

[0135] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for intelligent irrigation monitoring and control of paddy fields based on multi-source information fusion, characterized in that, Includes the following steps: Based on the multi-dimensional information collected in the early stage, a quantitative relationship model between fertilizer and water input and crop growth response was constructed using deep belief networks. Periodically collect real-time multi-source monitoring data of the planting area and determine the growth stage of rice seedlings; Based on the growth stages of rice seedlings, the target fertilizer and water requirements are determined through a quantitative relationship model between fertilizer and water input and crop growth response. Based on the improved DS evidence theory algorithm, real-time multi-source monitoring data is integrated to dynamically assess the environmental status of the current planting area according to the target fertilizer and water requirements, and combined with future weather forecast data to generate irrigation decision results that take into account both crop needs and resource conservation. Based on the fuzzy PID control algorithm, the fertilizer and water irrigation operation is automatically controlled according to the irrigation decision results; Based on an improved DS evidence theory algorithm, this method integrates real-time multi-source monitoring data, dynamically assesses the current state of the planting area according to the target fertilizer and water requirements, and combines future weather forecast data to generate irrigation decision results that balance crop needs and resource conservation. The specific steps include: Define a framework for identifying the state of planting areas; The basic weights are dynamically assigned based on the preset sensor type and environmental conditions, and the weight coefficients are dynamically adjusted based on future weather forecast data to obtain the dynamic weights of each type of data in the multi-source monitoring data. Calculate the basic probability allocation function for each type of data based on the dynamic weights of each type of data; Based on the basic probability allocation function of each type of data, real-time multi-source monitoring data are integrated to determine the state confidence level; When conflicts occur among different types of data in multi-source monitoring data, the entropy weight method is used to correct the dynamic weights. Based on the state confidence level, determine the water demand index and It then uses preset thresholds to make judgments and generate irrigation decision results; The fuzzy PID control algorithm includes a dual closed-loop control structure: its outer loop is the target flow rate, which is calculated from the target fertilizer and water demand; the inner loop is for real-time adjustment of the solenoid valve opening and the fertilizer pump speed; the variable input expression of the fuzzy PID control algorithm is: ； Where t is time; e(t) is the flow deviation; ec(t) is the rate of change of deviation; Q set For target traffic; Q actual The actual flow rate is used; the PID parameters are dynamically adjusted based on the fuzzy results of the flow rate deviation and the rate of change of deviation.

2. The method for intelligent irrigation and control of paddy fields based on multi-source information fusion according to claim 1, characterized in that, The multi-dimensional information includes soil moisture data, meteorological environmental parameters, and crop growth status; the expression for the quantitative relationship model between fertilizer and water input and crop growth response is as follows: ； Where Y is the output crop response variable; X i is the input multi-dimensional information vector; W is the weight matrix, optimized through backpropagation algorithm; b is the bias term.

3. The intelligent irrigation and control method for paddy fields based on multi-source information fusion according to claim 1, characterized in that, The multi-source monitoring data is collected periodically through soil parameter monitoring units, meteorological environment monitoring units, and crop growth monitoring units, based on a sensor network deployed in the planting area.

4. The method for intelligent irrigation and control of paddy fields based on multi-source information fusion according to claim 1, characterized in that, The rules for generating irrigation decision results are as follows: Irrigation should be initiated when the water demand index = bel (water shortage) + 0.5 * Pl (water shortage) > 0.7; When the fertilizer requirement index equals bel (nitrogen deficiency) Fertilization is triggered when Pl (phosphorus deficiency) > 0.6; where bel is the trust function and Pl is the likelihood function.

5. The method for intelligent irrigation and control of paddy fields based on multi-source information fusion according to claim 1, characterized in that, The mapping relationship output by the fuzzy PID control algorithm is as follows: ； Among them, K v N represents the opening degree of the solenoid valve. p U(t) represents the rotational speed of the fertilizer pump; u(t) represents the final control variable; the final control variable is calculated using the center of gravity method. ； Where, k p k i k d These are the PID parameters, representing the proportional coefficient, integral coefficient, and derivative coefficient, respectively.

6. The method for intelligent irrigation and control of paddy fields based on multi-source information fusion according to any one of claims 1-5, characterized in that, It also includes the following steps: Remotely monitor fertigation operations and optimize the model based on actual irrigation execution parameters and effect data.

7. A smart irrigation and fertilization control system for paddy fields based on multi-source information fusion, used to implement the smart irrigation and fertilization control method for paddy fields as described in any one of claims 1-6, characterized in that, It includes a multi-source information acquisition system, a computer system, and a mechanical execution system; the multi-source information acquisition system is used to periodically collect real-time multi-source monitoring data of the planting area and determine the growth stage of rice seedlings; The computer system includes an edge computing gateway and an intelligent control system; The edge computing gateway includes: The model building module is used to construct a quantitative relationship model between fertilizer and water input and crop growth response based on multi-dimensional information collected in the early stage and deep belief network. The target input determination module is used to determine the target fertilizer and water requirements based on the growth stage of rice seedlings through a quantitative relationship model between fertilizer and water input and crop growth response. Data fusion and irrigation decision module: Based on the improved DS evidence theory algorithm, it integrates real-time multi-source monitoring data, dynamically assesses the environmental status of the current planting area according to the target fertilizer and water requirements, and combines future weather forecast data to generate irrigation decision results that take into account both crop needs and resource conservation. The intelligent control system includes: The automatic control module is used to automatically control the fertigation operation based on the irrigation decision results using a fuzzy PID control algorithm.

8. The intelligent irrigation and control system for paddy fields based on multi-source information fusion according to claim 7, characterized in that, Also includes: The cloud-based remote monitoring and optimization system is used to remotely monitor fertigation operations and optimize the model based on actual irrigation execution parameters and effect data.

Images