A method of sprinkler control by communication link attenuation inversion

Abstract

This invention discloses a sprinkler irrigation control method based on communication link attenuation inversion, belonging to the fields of agricultural automatic sprinkler irrigation technology and wireless communication engineering technology. The method includes: deploying multiple fixed communication nodes in farmland to form a mesh link; each node sequentially sending ranging signals and recording the actual received power of each link; calculating the theoretical power using a free-space loss model to obtain the additional attenuation; discretizing the farmland into a three-dimensional voxel grid and establishing a system of linear equations relating the attenuation to the voxel attenuation coefficient; solving the equations using an algebraic iterative reconstruction algorithm to obtain the soil moisture distribution matrix within the farmland; identifying regions based on the moisture distribution and generating differentiated sprinkler irrigation commands; and controlling the electromagnetic valves of each zone via a wireless network to perform variable-volume sprinkler irrigation. This invention eliminates the need for buried soil moisture sensors, utilizes communication link tomography technology to obtain a true three-dimensional moisture field, is unaffected by weather, achieves precise closed-loop sprinkler irrigation, and significantly improves water-saving efficiency and crop uniformity.

Description

Technical Field

[0001] This invention relates to the fields of agricultural automatic sprinkler irrigation technology and wireless communication engineering technology, and in particular to a sprinkler irrigation control method based on communication link attenuation inversion. Background Technology

[0002] In existing technologies, automatic sprinkler irrigation control in agriculture typically employs embedded soil moisture sensors for moisture detection. This method involves deploying several point-type moisture sensors (such as frequency domain reflectance or time domain reflectance sensors) in the farmland to directly measure the soil moisture content within a very small area around the sensors. Sprinkler irrigation is activated when the measured value falls below a threshold. However, soil moisture exhibits high spatial heterogeneity, and the sensors can only obtain information from discrete points. The limited number of sensors cannot reflect the true distribution of water shortage across the entire farmland. Furthermore, long-term burial of sensors in moist, acidic, alkaline, or saline soil environments can lead to electrochemical corrosion of the electrodes, causing measurement drift, poor long-term system reliability, and high maintenance costs.

[0003] Existing technologies also employ sprinkler irrigation control methods based on meteorological parameters and crop evapotranspiration models. This method collects parameters such as temperature, humidity, wind speed, and solar radiation from weather stations, uses empirical formulas to estimate crop evapotranspiration, and then performs timed and quantitative sprinkler irrigation. However, this approach ignores local differences within the field (such as uneven soil moisture caused by varying elevations and textures), resulting in significant water waste and potential for over-irrigation in some areas and under-irrigation in others, leading to uneven crop growth.

[0004] Existing technologies also employ spectral remote sensing or UAV multispectral imaging for water monitoring. This method calculates the water stress index using the spectral reflectance characteristics of the crop canopy, but it can only acquire information from the canopy surface and cannot penetrate the canopy to perceive the distribution of deep soil moisture. Furthermore, it is heavily dependent on sunny, cloudless lighting conditions and cannot function on cloudy or rainy days, which are precisely the critical periods for sprinkler irrigation decisions.

[0005] Existing technologies also attempt to infer soil moisture using the signal attenuation of a single or a few wireless links. This method relies on the correlation between electromagnetic wave attenuation in soil and the dielectric constant (and consequently, water content), using the attenuation of one or a few links to infer the average soil moisture along the path. However, these schemes only obtain an overall average value along the path, failing to distinguish the specific spatial distribution of moisture, let alone construct two-dimensional or three-dimensional moisture field maps. When localized drought and localized waterlogging coexist in farmland, the average attenuation of a single link may appear "normal," thus masking the true spatial heterogeneity.

[0006] In summary, existing technologies generally suffer from problems such as reliance on discrete point sensors, significant susceptibility to weather conditions, the ability to acquire only surface information, or the inability to obtain path averaging results without imaging. To date, no research has proposed a large-scale, gridded deployment of communication nodes in farmland to use signal attenuation from all links as projection data, obtaining three-dimensional soil moisture distribution through tomographic imaging inversion, and then using this data for closed-loop control of differentiated sprinkler irrigation operations. This invention addresses these technological gaps and deficiencies. Summary of the Invention

[0007] The purpose of this invention is to provide a sprinkler irrigation control method based on communication link attenuation inversion, which solves the problems of existing technologies that rely on discrete point sensors, cannot obtain three-dimensional water distribution, are greatly affected by weather, and have extensive sprinkler irrigation control.

[0008] To achieve the above objectives, the present invention provides a sprinkler irrigation control method based on communication link attenuation inversion, comprising the following steps: Step S1: Deploy multiple fixed communication nodes in the target farmland area. Each communication node is capable of transmitting and receiving wireless ranging signals, forming multiple mesh communication links between all communication nodes. Step S2: Control all communication nodes to sequentially send high-frequency continuous wave ranging signals and record the actual power value of the received signal on each communication link. ,in and These are the numbers used to designate different communication nodes; Step S3: Based on the actual length of each communication link By combining the free-space path loss model, the theoretical received power value of each communication link under dry air conditions is calculated. ; Step S4: Compare the actual received power with the theoretical received power for each communication link, and calculate the additional attenuation for each link. ; Step S5: Discretize the target farmland area into a three-dimensional voxel grid, with each voxel corresponding to an unknown soil moisture dielectric constant, and construct a system of linear equations between the additional attenuation and the three-dimensional voxel grid: ; in, For communication link - The total number of voxels, For the first The decay coefficient per unit length of a voxel. For communication links - Passing through the The length of a voxel; Step S6: Solve the linear equation system using an algebraic iterative reconstruction algorithm to obtain the attenuation coefficient distribution of all voxels, and then determine the calibration relationship between the attenuation coefficient and the soil volumetric water content. The three-dimensional soil moisture distribution matrix inside the farmland was obtained by inversion; Step S7: Based on the three-dimensional soil moisture distribution matrix, identify water-deficient areas and areas with excessive moisture content, and generate differentiated sprinkler irrigation instructions; Step S8: Send the sprinkler irrigation command to the solenoid valve control nodes of each zone in the farmland via wireless network, and control the sprinklers in the corresponding areas to perform variable water volume sprinkler irrigation.

[0009] Preferably, the multiple fixed communication nodes deployed in step S1 adopt a honeycomb or triangular grid topology, with a horizontal spacing of 5 to 20 meters between adjacent nodes, and a layer of communication nodes deployed vertically at 10 centimeters below the soil surface and 50 centimeters above the ground surface.

[0010] Preferably, the carrier frequency of the high-frequency continuous wave ranging signal in step S2 is 2.4 GHz or 5.8 GHz, the signal transmission power is dynamically adjustable between 10 dBm and 20 dBm, and each communication node uses time division multiple access to send ranging signals sequentially to avoid co-channel interference.

[0011] Preferably, the specific expression for the free space path loss model in step S3 is as follows: ; in, For transmission power, and These are the transmit antenna gain and the receive antenna gain, respectively. For signal frequency, The unit is kilometers.

[0012] Preferably, when constructing the linear equation system in step S5, a regularization constraint term is also introduced. To suppress ill-conditioned matrix problems in the solution process, where This is the Tikhonov regularization parameter, with a value ranging from 0.01 to 0.1.

[0013] Preferably, the algebraic iterative reconstruction algorithm in step S6 is a joint iterative reconstruction algorithm, in which the decay coefficients of all voxels are updated synchronously during each iteration, and the iteration termination condition is set to the root mean square error of soil moisture distribution reconstructed in two consecutive iterations being less than 0.5%.

[0014] Preferably, the specific method for identifying water-deficient areas in step S7 is as follows: the three-dimensional soil moisture distribution matrix is ​​compared with the preset upper and lower limits of suitable moisture content for each growth stage of crops. When the volume moisture content of a certain voxel is lower than the lower limit threshold by more than 10%, it is determined to be a water-deficient area; when the volume moisture content of a certain voxel is higher than the upper limit threshold by more than 15%, it is determined to be an area with excessive moisture content.

[0015] Preferably, in step S8, the variable water volume sprinkler operation uses pulse width modulation to control the opening duty cycle of the solenoid valve. The duty cycle of the sprinkler corresponding to the water shortage area is set to 80% to 100%, the duty cycle of the sprinkler corresponding to the water content exceeding the standard is set to 0% to 20%, and the duty cycle of the sprinkler corresponding to the normal area is set to 30% to 70%.

[0016] Preferably, a system self-calibration step is included before step S2: after a one-time full irrigation to bring the soil moisture to field capacity, a complete attenuation measurement process is performed, and the reference attenuation value of each link is recorded at this time. The reference attenuation value is used as the zero-point offset correction parameter for subsequent inversion calculation.

[0017] Preferably, it also includes a dynamic update step: repeating steps S2 to S6 in a time period of 10 to 30 minutes to update the three-dimensional soil moisture distribution matrix in real time, and dynamically adjusting the sprinkler irrigation command based on the updated matrix to form a closed-loop control.

[0018] Therefore, the sprinkler control method based on communication link attenuation inversion using the above-described structure has the following beneficial effects: (1) This invention does not rely on any buried soil moisture sensor, thus completely avoiding sensor corrosion, degradation and maintenance problems; and by utilizing the volume penetration characteristics of communication signals in the soil, it can invert the true three-dimensional moisture distribution, rather than just the information of the surface or a certain depth point.

[0019] (2) Based on the physical principle of electromagnetic wave attenuation, this invention is not affected by weather conditions such as rain or sunshine. The method is indirectly measured by tomographic imaging technology, with very few comparative documents, and has outstanding substantive features and non-obviousness. The closed-loop control strategy in this invention significantly saves irrigation water, and the expected water saving rate can reach more than 30%.

[0020] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the farmland communication node deployment and link in an embodiment of the present invention; Figure 2 This is an overall flowchart of the sprinkler irrigation control method of the present invention; Figure 3 This is a schematic diagram illustrating the three-dimensional voxel mesh division and the geometric relationship of a single link passing through a voxel in an embodiment of the present invention; Figure 4 This is an example of a three-dimensional color map of soil moisture distribution obtained by inversion in an embodiment of the present invention; Figure 5 This is a schematic diagram of the differentiated sprinkler irrigation zoning instructions generated based on water distribution in an embodiment of the present invention. Detailed Implementation

[0022] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0023] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0024] Example 1 (Application in Field Grain Crops) This embodiment uses a winter wheat field with an area of ​​1 hectare (100 meters × 100 meters) as an example for illustration. See [link to example]. Figure 1-5 .

[0025] I. System Deployment like Figure 1 As shown, 36 fixed communication nodes were deployed within and along the boundaries of the farmland. They were arranged in a 6x6 grid on the horizontal plane, with a horizontal spacing of 20 meters between adjacent nodes. Each node was capable of transmitting and receiving wireless signals and operated at 2.4... Frequency band, transmit power 15 Antenna gain 2 All nodes form communication links with each other, totaling 630 mesh links.

[0026] Two layers of nodes are installed vertically: the first layer is buried 10 cm below the ground surface to sense soil moisture in the crop root zone (0-30 cm); the second layer is installed on a pole 50 cm above the ground surface to correct for the influence of the near-surface atmospheric environment on signal propagation. The nodes are powered by solar cells and equipped with ZigBee modules to aggregate data to a central controller at the field edge.

[0027] II. System Self-Calibration After the system's initial operation or after each full irrigation to bring the soil to field capacity, a self-calibration step is performed. The central controller controls all nodes to sequentially send ranging signals, recording the received power of each link as a reference value for subsequent zero-point offset correction.

[0028] III. Signal Acquisition and Calculation of Additional Attenuation Reference Figure 2 The process shown involves the central controller using time-division multiple access (TDMA) to sequentially trigger each node to transmit continuous wave ranging signals. Each time slot is 10 milliseconds, and a complete cycle is 360 milliseconds. For each node... Launch, Node The receiving link records the actual received power. .

[0029] With nodes and For example, the distance between the two is 20 meters. The theoretical received power is calculated based on the free-space path loss model: ; in , , Substitute the values: ; ; Actual measurement Then the additional attenuation is: ; Calculate the values ​​for each of the 630 links one by one to obtain the value for each link. .

[0030] IV. Three-dimensional voxel mesh generation and construction of linear equations like Figure 3 As shown, the farmland is discretized into a three-dimensional voxel grid. Horizontally, there are 10×10 voxels, each measuring 10 meters × 10 meters; vertically, there are two layers, corresponding to the 0–30 cm root layer and the 30–60 cm depth layer, totaling 200 voxels. Each voxel corresponds to the attenuation coefficient per unit length to be determined. (Unit: dB / m).

[0031] For any link Calculate the path length through each voxel. For example, if a link passes through 10 voxels in sequence, with travel lengths of 3.2 m, 4.1 m, ..., 2.8 m, the equation can be obtained as follows: ; Equations are listed for all links, forming a system of linear equations: ; in, For link The total number of voxels traversed. To suppress the ill-conditioned matrix problem, a Tikhonov regularization term is introduced. Regularization parameters .

[0032] V. Solving the Algebraic Iterative Reconstruction Algorithm A joint iterative reconstruction algorithm is used to solve the problem. Initially, the attenuation coefficients of all voxels are set. The iteration step size is 0.02. During each iteration, the equations of all links are used to synchronously update the value of each voxel. The iteration termination condition was set to the root mean square error of the soil moisture distribution reconstructed between two consecutive iterations being less than 0.5%. In this example, the convergence condition was met after 15 iterations.

[0033] VI. Inversion of 3D Soil Moisture Distribution Based on the pre-calibrated relationship between the attenuation coefficient and soil volumetric water content in the laboratory, an exponential model is adopted: ; in Units are , This is the volumetric water content (decimal form). For a given voxel... The calculation yields: ; Calculations were performed on all 200 voxels individually, resulting in the following: Figure 4 The three-dimensional soil moisture distribution matrix is ​​shown (blue represents low water content areas and red represents high water content areas in the map).

[0034] VII. Identifying Water-Scarce Areas and Generating Differentiated Sprinkler Irrigation Instructions For winter wheat during the jointing stage, the optimal moisture content is preset to a lower limit of 15% and an upper limit of 20%. The judgment rules are as follows: when the volumetric moisture content of a certain voxel is more than 10% lower than the lower limit threshold (i.e., below 13.5%), it is judged as a water-deficient area; when it is more than 15% higher than the upper limit threshold (i.e., above 23%), it is judged as an area with excessive moisture content.

[0035] The inversion results of this embodiment show that the moisture content of the 8 voxels in the northwest corner of the farmland (corresponding to the two layers of voxels in the 1st and 2nd rows and the 1st and 2nd columns) is only 8% to 11%, which is identified as a water-deficient area; the moisture content of the 6 voxels in the southeast corner (corresponding to the lower layers of voxels in the 9th and 10th rows and the 9th and 10th columns) is 26% to 30%, which is identified as an area with excessive moisture content; the rest are normal areas.

[0036] like Figure 5 As shown, the central controller generates differentiated irrigation commands: the duty cycle of the solenoid valve in the water-scarce area is set to 90%, the duty cycle in the area with excessive water content is set to 0% (no irrigation), and the duty cycle in the normal area is set to 50%.

[0037] 8. Perform sprinkler irrigation operations The central controller sends pulse-width modulation (PWM) irrigation commands to the solenoid valves of each zone via a ZigBee wireless network. The solenoid valves in water-scarce areas spray continuously at a 90% duty cycle for 30 minutes and then close; those in normal areas spray at a 50% duty cycle for 15 minutes and then close; and those in areas with excessive water content remain closed.

[0038] IX. Dynamic Updates and Closed-Loop Control The system repeats steps three through seven in a 20-minute cycle, updating the three-dimensional soil moisture distribution matrix in real time. After each update, the central controller regenerates irrigation commands based on the latest moisture distribution and issues them for execution, forming a closed-loop adaptive control. Comparative experiments show that this embodiment saves 32% of water compared to traditional timed and quantitative sprinkler irrigation, and significantly improves crop growth uniformity.

[0039] Example 2 (Application in Slope Tea Gardens) This embodiment uses a gently sloping tea garden with an area of ​​0.5 hectares (50 meters × 100 meters) and an average slope of 15 degrees as an example to illustrate the application of this invention in complex terrain and shallow-rooted crops. The overall process is still referred to Figure 2 For the geometric relationship between voxel partitioning and link crossings, see [link to related information]. Figure 3 See the moisture distribution map. Figure 4 For partition control, see [link / reference] Figure 5 .

[0040] I. System Deployment For sloping terrain, a non-uniform triangular mesh topology was used to deploy communication nodes. Nodes were densified at the edges of terraces and in areas with drastic slope changes, with horizontal spacing ranging from 5 to 15 meters, for a total of 28 nodes, forming 378 communication links.

[0041] Vertically, tea tree roots are mainly distributed in the 0-40 cm soil layer, therefore only two layers of monitoring points are set: the first layer is buried 5 cm below the surface (to monitor surface moisture), and the second layer is buried 25 cm below the surface (to monitor moisture in the main root layer). No above-ground monitoring points are set. The operating frequency is selected as 5.8. This frequency band is more sensitive to changes in shallow water content. The transmit power is dynamically adjusted to 5. Up to 15 This is to reduce long-distance interference of high-frequency signals.

[0042] II. System Self-Calibration Because slopes are prone to runoff, leading to drastic changes in water distribution, the frequency of the self-calibration step was increased to once per hour. A complete attenuation measurement was performed after a single, thorough irrigation, and the baseline attenuation value for each link was recorded as a zero-point offset correction parameter.

[0043] III. Signal Acquisition and Calculation of Additional Attenuation The received power of all links is collected using time-division multiple access (TDMA). Taking a typical link as an example: Node... and The actual distance between them is 12 meters. ,frequency Transmission power Antenna gain Theoretical received power: ; Calculated , ,therefore: ; Actual measurement Then the additional attenuation is: ; Calculate each of the 378 links individually.

[0044] IV. Three-dimensional voxel mesh generation and construction of linear equations The horizontal axis is divided into 5m x 5m voxels (10 rows x 20 columns = 200 voxels), and the vertical axis is divided into two layers (0~15 cm and 15~40 cm), for a total of 400 voxels. The number of links (378) is slightly less than the number of voxels, but this can be mitigated by introducing regularization constraints. By utilizing the prior knowledge of the smoothness of water distribution, a stable solution can still be obtained. For each link, the path length passing through each voxel is calculated. List in the form of The equation.

[0045] V. Solving the Algebraic Iterative Reconstruction Algorithm A joint iterative reconstruction algorithm is adopted, with initial settings... The iteration step size was 0.015. Since the equations were slightly underdetermined, the number of iterations was increased and stricter convergence conditions were set (the root mean square error of the soil moisture distribution between the two reconstructions was less than 0.3%). This example converged after 22 iterations.

[0046] VI. Inversion of 3D Soil Moisture Distribution A power function relationship was pre-calibrated for the tea garden soil: ; in Units are , This is the volumetric water content (in decimal form). For example, a certain voxel... ,but .

[0047] calculate ; have to It is located in an arid region.

[0048] VII. Identifying Water-Scarce Areas and Generating Differentiated Sprinkler Irrigation Instructions A precise assessment is made for different degrees of water shortage in tea trees: when the volumetric moisture content is 5% to 10% below the suitable lower limit (set to 10%), it is judged as mild water shortage, and the nozzle duty cycle is set to 40% to 60%; when it is 10% to 20% below the lower limit, it is judged as moderate water shortage, and the duty cycle is set to 60% to 80%; when it is more than 20% below the lower limit, it is judged as severe water shortage, the duty cycle is set to 100%, and the watering cycle is extended by one time; when it is more than 15% above the upper limit (set to 20%), it is judged as excessive moisture content, and the duty cycle is set to 0%.

[0049] The inversion results of this embodiment show that approximately 24 voxel moisture contents in the upper slope area of ​​the tea garden are below 6%, indicating severe water shortage; while approximately 12 voxel moisture contents in the low-lying area at the foot of the slope are above 25%, indicating excessive water content. The central controller generates differentiated sprinkler irrigation instructions based on this information.

[0050] 8. Perform sprinkler irrigation operations The tea garden uses a drip irrigation system, with each voxel corresponding to a solenoid valve on a drip irrigation branch pipe. The central controller sends pulse width modulation commands via a wireless network: severely water-deficient areas are continuously drip-irrigated at 100% duty cycle for 45 minutes, mildly water-deficient areas are drip-irrigated at 50% duty cycle for 20 minutes, and valves are shut off in areas with excessive moisture content.

[0051] IX. Dynamic Updates and Closed-Loop Control The data was collected and analyzed repeatedly in 15-minute intervals. After one week of system operation, statistics showed that water was saved by 28% compared to traditional manual irrigation, the uniformity of tea shoot sprouting was improved, and the quality indicators of tea (total free amino acids) increased by about 12%.

[0052] Example 3 (Application in greenhouse agriculture) This embodiment uses a 2000-square-meter (40m x 50m) multi-span tomato greenhouse as an example to illustrate the intelligent irrigation solution integrating the present invention with a mobile robot. The overall process of this embodiment is still as described above. Figure 2 The voxel division is more refined; see the water distribution map for details. Figure 4 For partition control, see [link / reference] Figure 5 .

[0053] I. System Deployment This embodiment employs a hybrid architecture combining static nodes and dynamic mobile nodes. Static backbone nodes: Eight communication nodes are fixedly installed at the four corners and the central support pillar of the greenhouse, 1.5 meters above the ground, serving as network references and clock synchronization sources. Dynamic mobile nodes: Communication nodes are added to the existing track-mounted inspection robot in the greenhouse. When performing disease inspection tasks, the robot moves at a constant speed (0.2 m / s) along a preset path. Its communication nodes interact with the static nodes in real time, forming time-varying, high-density virtual links. For example, a single static node and a mobile node can form countless links traversing different spaces at different times, greatly enriching the amount of projected data. Approximately 3200 effective links are formed between the mobile nodes and static nodes (within a single inspection cycle).

[0054] No soil nodes are buried in the vertical direction. The antenna of the mobile node is 0.3 meters above the ground (close to the crop canopy), and the antenna of the static node is 1.5 meters above the ground.

[0055] II. System Self-Calibration Self-calibration is performed when the greenhouse is fully irrigated and drained to field capacity for the first time after transplanting. The mobile node travels along a fixed route at a low speed (0.1 m / s) and records the received power of each link (including static static and static moving links) as a reference value.

[0056] III. Signal Acquisition and Calculation of Additional Attenuation A time-division multiple access (TDMA) approach is employed, prioritizing communication between static nodes and between static nodes and mobile nodes. The mobile node transmits a ranging signal every 0.5 meters of movement, using a frequency of 2.4 GHz. Transmission power 10 Antenna gain 2 .

[0057] For a static mobile link, let the static node be... With mobile nodes At a certain moment, the straight-line distance is 15 meters (0.015 km), and the theoretical receiving power is: ; Calculated , ,but: ; Actual measurement Additional attenuation: ; Calculate each link individually.

[0058] IV. Three-dimensional voxel mesh generation and construction of linear equations The focus was on the densely populated tomato root zone, from 0 to 60 cm. A 1m x 1m ultra-high-resolution grid was created horizontally (40 x 50 = 2000 voxels), and six vertical layers were created, spaced 10 cm apart, for a total of 12000 voxels. Each voxel corresponds to a unit length attenuation coefficient. .

[0059] Because mobile nodes bring massive amounts of spatiotemporally correlated link data, compressed sensing reconstruction algorithms are used to replace traditional algebraic iterative reconstruction. The problem is formulated as follows: ; in It is a coefficient matrix (link path length matrix). To add an additional attenuation vector to the actual measurement, Let be the vector of attenuation coefficients to be determined. This is the allowable error.

[0060] V. Solving with Compressed Sensing Algorithm The above optimization problem is solved using a basis pursuit denoising algorithm, taking advantage of the prior knowledge of the spatial sparsity of water distribution (i.e., water content changes gradually in most areas, with a few areas exhibiting abrupt changes). Iterative calculations are performed until the residual is less than a set threshold, at which point the output is obtained. The sparse solution is obtained. In this embodiment, the number of iterations is approximately 80, and a single reconstruction takes approximately 2 seconds on a central controller configured with a GPU.

[0061] VI. Inversion of 3D Soil Moisture Distribution The calibration relationship uses a linear exponential piecewise model fitted to greenhouse sandy loam soil: ; in Units are , This represents the volumetric water content (decimal form). Calculations were performed on each of the 12,000 voxels, yielding the following results: Figure 4 The high-resolution three-dimensional moisture distribution map shown is shown.

[0062] VII. Identifying Water-Scarce Areas and Generating Differentiated Sprinkler Irrigation Instructions For tomatoes during the fruiting stage, the suitable moisture content was set at a lower limit of 12% and an upper limit of 18%. A voxel with a moisture content below 10% was identified as a water-deficient area, and above 21% was identified as an area with excessive moisture. The results of this embodiment showed that at the greenhouse entrance, due to frequent door opening and closing leading to strong ventilation and high evaporation, approximately 150 voxels had a moisture content below 9%; in three sections at the far end of the greenhouse, due to drip irrigation pipe blockage, approximately 80 voxels had a moisture content below 8%. The central controller generated instructions: a 100% duty cycle (emergency water replenishment) for water-deficient areas, and simultaneously sent a blockage alarm to management personnel; a 0% duty cycle was set for areas with excessive moisture (approximately 40 voxels, due to a malfunctioning drip irrigation valve that remained open); and a 40% duty cycle for normal areas.

[0063] 8. Implementing sprinkler irrigation operations and multi-factor linkage The greenhouse uses a drip irrigation system, and the central controller sends pulse width modulation commands to each solenoid valve via a LoRa wireless network. Simultaneously, this embodiment uploads a three-dimensional soil moisture distribution matrix to the greenhouse environmental control system in real time. When a region is identified as experiencing persistent water shortage due to a sprinkler system malfunction, the system automatically lowers the set temperature in that region (by 2°C) to reduce crop transpiration and prevent severe drought stress.

[0064] IX. Dynamic Updates and Closed-Loop Control The mobile robot completes a full greenhouse inspection every 10 minutes, updating the water distribution matrix and adjusting the sprinkler commands after each inspection. After one month of operation, statistics show that the system saves 35% of water compared to traditional timed drip irrigation, improves the uniformity of tomato fruit size by 18%, and reduces the incidence of hollow fruit to below 5%.

[0065] Therefore, the sprinkler irrigation control method based on communication link attenuation inversion of the above structure can obtain the true three-dimensional soil moisture distribution without relying on embedded sensors, and implement precise variable water volume sprinkler irrigation accordingly, which has significant water-saving and efficiency-enhancing effects.

[0066] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A sprinkler irrigation control method based on communication link attenuation inversion, characterized in that, Includes the following steps: Step S1: Deploy multiple fixed communication nodes in the target farmland area. Each communication node is capable of transmitting and receiving wireless ranging signals, forming multiple mesh communication links between all communication nodes. Step S2: Control all communication nodes to sequentially send high-frequency continuous wave ranging signals and record the actual power value of the received signal on each communication link. ,in and These are the numbers used to designate different communication nodes; Step S3: Based on the actual length of each communication link By combining the free-space path loss model, the theoretical received power value of each communication link under dry air conditions is calculated. ; Step S4: Compare the actual received power with the theoretical received power for each communication link, and calculate the additional attenuation for each link. ; Step S5: Discretize the target farmland area into a three-dimensional voxel grid, with each voxel corresponding to an unknown soil moisture dielectric constant, and construct a system of linear equations between the additional attenuation and the three-dimensional voxel grid: ； in, For communication link - The total number of voxels, For the first The decay coefficient per unit length of a voxel. For communication links - Passing through the The length of the voxel; Step S6: Solve the linear equation system using an algebraic iterative reconstruction algorithm to obtain the attenuation coefficient distribution of all voxels, and then determine the calibration relationship between the attenuation coefficient and the soil volumetric water content. The three-dimensional soil moisture distribution matrix inside the farmland was obtained by inversion; Step S7: Based on the three-dimensional soil moisture distribution matrix, identify water-deficient areas and areas with excessive moisture content, and generate differentiated sprinkler irrigation instructions; Step S8: Send the sprinkler irrigation command to the solenoid valve control nodes of each zone in the farmland via wireless network, and control the sprinklers in the corresponding areas to perform variable water volume sprinkler irrigation.

2. The sprinkler control method based on communication link attenuation inversion according to claim 1, characterized in that, The multiple fixed communication nodes deployed in step S1 adopt a honeycomb or triangular grid topology. The horizontal spacing between adjacent nodes is 5 to 20 meters, and a layer of communication nodes is deployed vertically at 10 centimeters below the soil surface and 50 centimeters above the ground surface.

3. The sprinkler control method based on communication link attenuation inversion according to claim 1, characterized in that, In step S2, the carrier frequency of the high-frequency continuous wave ranging signal is 2.4 GHz or 5.8 GHz, the signal transmission power is dynamically adjustable between 10 dBm and 20 dBm, and each communication node uses time division multiple access to send ranging signals sequentially to avoid co-channel interference.

4. The sprinkler control method based on communication link attenuation inversion according to claim 1, characterized in that, The specific expression for the free space path loss model in step S3 is as follows: ； in, For transmission power, and These are the transmit antenna gain and the receive antenna gain, respectively. For signal frequency, The unit is kilometers.

5. The sprinkler control method based on communication link attenuation inversion according to claim 1, characterized in that, In step S5, when constructing the system of linear equations, a regularization constraint term is also introduced. To suppress ill-conditioned matrix problems in the solution process, where This is the Tikhonov regularization parameter, with a value ranging from 0.01 to 0.

1.

6. The sprinkler control method based on communication link attenuation inversion according to claim 1, characterized in that, In step S6, the algebraic iterative reconstruction algorithm is a joint iterative reconstruction algorithm. During each iteration, the decay coefficient of all voxels is updated synchronously. The iteration termination condition is set to the root mean square error of soil moisture distribution reconstructed in the previous two iterations being less than 0.5%.

7. The sprinkler control method based on communication link attenuation inversion according to claim 1, characterized in that, The specific method for identifying water-deficient areas in step S7 is as follows: compare the three-dimensional soil moisture distribution matrix with the preset upper and lower limits of suitable moisture content for each growth stage of crops. When the volume moisture content of a certain voxel is more than 10% lower than the lower limit, it is determined to be a water-deficient area; when the volume moisture content of a certain voxel is more than 15% higher than the upper limit, it is determined to be an area with excessive moisture content.

8. The sprinkler control method based on communication link attenuation inversion according to claim 1, characterized in that, In step S8, the variable water volume sprinkler operation uses pulse width modulation to control the opening duty cycle of the solenoid valve. The duty cycle of the sprinkler corresponding to the water shortage area is set to 80% to 100%, the duty cycle of the sprinkler corresponding to the water content exceeding the standard is set to 0% to 20%, and the duty cycle of the sprinkler corresponding to the normal area is set to 30% to 70%.

9. The sprinkler irrigation control method based on communication link attenuation inversion according to any one of claims 1 to 8, characterized in that, Before step S2, a system self-calibration step is also included: after a one-time full irrigation to bring the soil moisture to field capacity, a complete attenuation measurement process is performed, and the reference attenuation value of each link is recorded at this time. This reference attenuation value is used as the zero-point offset correction parameter for subsequent inversion calculation.

10. The sprinkler irrigation control method based on communication link attenuation inversion according to any one of claims 1 to 8, characterized in that, It also includes a dynamic update step: with a time period of 10 to 30 minutes, steps S2 to S6 are repeated to update the three-dimensional soil moisture distribution matrix in real time, and the sprinkler irrigation command is dynamically adjusted based on the updated matrix to form a closed-loop control.

Images