Environment data wireless monitoring method and system for smart agriculture

Abstract

The invention provides an environmental data wireless monitoring method and system for smart agriculture, and relates to the technical field of wireless monitoring, and the method comprises the steps: respectively deploying a LoRa transmitting node and a receiving node at two sides of a to-be-monitored soil region; controlling the transmitting node to transmit wireless signals on the first working frequency band and the second working frequency band respectively; the receiving node receives and records the first receiving signal strength and the second receiving signal strength; acquiring a first baseline signal strength and a second baseline signal strength; calculating to obtain a first frequency band attenuation variable quantity and a second frequency band attenuation variable quantity; calculating to obtain a differential attenuation variable; substituting the differential attenuation variable into the soil humidity inversion model, and calculating to obtain the volume water content VWC value of the current soil; and performing judgment and early warning based on the volumetric water content VWC value or the change rate of the volumetric water content VWC value along with time. The method provided by the invention not only can realize non-contact and large-range soil humidity monitoring, but also can effectively improve the anti-interference performance and accuracy of the monitoring result.

Description

Technical Field

[0001] The present invention relates to the field of wireless monitoring technology, and in particular to a method for wireless monitoring of environmental data for smart agriculture. Background Art

[0002] Currently, for smart agriculture, soil moisture monitoring is generally based on point measurements using physical sensors. For example, commercial soil moisture sensors use frequency domain reflectometry or time domain reflectometry. These sensors need to directly contact the soil and accurately calculate the soil moisture at the sensor's buried point by measuring the soil's dielectric constant.

[0003] However, when applied to large areas of cultivated farmland, the deployment of soil moisture sensors will have the following defects: First, due to the large area of ​​cultivated farmland, a large number of physical sensors need to be deployed, and their hardware costs and maintenance costs are relatively high. In addition, there is also the risk of physical damage such as being crushed by agricultural machinery; second, physical sensors can only provide local soil moisture information at the installation point. If the overall humidity conditions of a large area of ​​cultivated farmland need to be obtained, a relatively dense deployment of sensor nodes is required, which is less economical. Summary of the Invention

[0004] In order to solve the technical problems in the related art, the present invention provides a wireless monitoring method for environmental data for smart agriculture.

[0005] In order to achieve the above object, the technical solution adopted by the present invention includes: According to a first aspect of the present invention, a method for wireless monitoring of environmental data for smart agriculture is provided, comprising the following steps: Step S1: A LoRa transmitter node and a LoRa receiver node are fixedly deployed on both sides of the soil area to be monitored; Step S2: Control the transmitting node to transmit wireless signals in a first working frequency band and a second working frequency band respectively, wherein the first working frequency band is different from the second working frequency band; Step S3: The receiving node receives and records the first received signal strength and the second received signal strength corresponding to the wireless signals in the first working frequency band and the second working frequency band; Step S4: obtaining a first baseline signal strength and a second baseline signal strength that were previously measured and stored under a soil baseline moisture state; and calculating a first frequency band attenuation change based on a difference between the first received signal strength and the first baseline signal strength, and calculating a second frequency band attenuation change based on a difference between the second received signal strength and the second baseline signal strength; Step S5: performing a linear combination operation on the attenuation variation of the first frequency band and the attenuation variation of the second frequency band to obtain a differential attenuation variable, so as to significantly suppress frequency-independent public environmental interference and partially suppress frequency-dependent environmental foreign body interference; Step S6: Substituting the differential attenuation variable into a pre-established soil moisture inversion model for characterizing the mapping relationship between the differential attenuation variable and the soil volumetric water content, and calculating the current soil volumetric water content VWC value; Step S7: making a judgment based on the volumetric water content VWC value or its rate of change over time, and generating a warning message when the judgment result exceeds a preset range.

[0006] Optionally, in step S2, the first operating frequency band and the second operating frequency band are selected from licensed frequency bands specified in the LoRa protocol, and the frequency difference between the two frequency bands is greater than 40 MHz.

[0007] Optionally, in step S4, the first baseline signal strength and the second baseline signal strength are obtained by controlling the transmitting node and the receiving node to perform signal measurement when the soil is at a known soil moisture after the node deployment is completed.

[0008] Optionally, in step S5, the linear combination operation specifically includes: The differential attenuation variable is calculated by subtracting the product of the calibration coefficient and the attenuation change of the second frequency band from the attenuation change of the first frequency band. : Where, is the attenuation change of the first frequency band, is the calibration coefficient, is the attenuation change of the second frequency band.

[0009] Optionally, the calibration coefficient is determined according to a set physical model, wherein the set physical model is used to characterize the functional relationship between the attenuation of the signal by environmental foreign matter and the signal frequency. The calculation expression is: Where, To set the functional relationship of the physical model, is the center frequency of the first operating frequency band, is the center frequency of the second operating frequency band.

[0010] Optionally, the functional relationship is modeled as the nth power function of the signal frequency, that is, The calculation expression is: Where n is the power exponent pre-calibrated through experiments.

[0011] Optionally, in step S6, the soil moisture inversion model is obtained by training a machine regression algorithm, and the data used to train the machine regression algorithm is the real soil moisture values ​​measured by traditional soil moisture sensors by synchronously collecting multiple groups of differential attenuation variables and their corresponding ones after the node is deployed.

[0012] Optionally, the machine learning regression algorithm is a support vector regression algorithm or a gradient boosting decision tree algorithm.

[0013] Optionally, the determination in step S7 specifically includes: Comparing the current volumetric water content (VWC) of the soil with a preset drought threshold, and generating a drought warning if the VWC is lower than the threshold; and / or, The rate of change of the current soil volumetric water content (VWC) per unit time is calculated. If the rate of change exceeds the preset water loss rate threshold or water filling rate threshold, a corresponding rapid water loss warning or rapid water filling warning is generated.

[0014] According to a second aspect of the present invention, there is further provided a wireless monitoring system for environmental data for smart agriculture, which is used to execute the wireless monitoring method for environmental data for smart agriculture described in any one of the technical solutions in the first aspect of the present invention. The wireless monitoring system for environmental data for smart agriculture comprises: Multiple pairs of fixedly deployed LoRa node groups, each pair of node groups corresponding to a monitoring sub-area, including a transmitting node and a receiving node; wherein the transmitting node is configured to respond to control instructions and alternately transmit wireless signals on a preset first working frequency band and a second working frequency band; the receiving node is configured to receive the wireless signals and measure the received signal strength on the first working frequency band and the second working frequency band; A central processing unit is communicatively connected to the receiving node and internally stores the first baseline signal strength, the second baseline signal strength, the calibration coefficient k, and the soil moisture inversion model; the central processing unit is configured to perform the following operations in sequence: receiving a first received signal strength and a second received signal strength from the receiving node; Calculating the attenuation variation of the first frequency band, the attenuation variation of the second frequency band, and the differential attenuation variable; The differential attenuation variable is input into the soil moisture inversion model to obtain the soil volume water content (VWC) value, and it is determined whether the VWC value or its rate of change exceeds a preset range. If so, an early warning module is triggered.

[0015] Beneficial effects: 1. Through the above technical solution, first, the method of the present invention can achieve non-contact, large-scale soil moisture monitoring. Specifically, the method of the present invention only requires the deployment of LoRa transmitting nodes and LoRa receiving nodes above the soil area to be measured, and the soil moisture is inverted by analyzing the transmission characteristics of the wireless signal between the nodes. There is no need to deploy physical sensors in the soil. This not only effectively avoids physical damage to the physical sensors caused by agricultural machinery rolling, soil corrosion or root entanglement, but also significantly reduces hardware costs and deployment and maintenance difficulties. In addition, the monitoring results obtained by a single pair of nodes are the comprehensive effect of the area below its signal propagation path. Compared with the point-type (buried point) measurement results of physical sensors, they are more representative of the overall humidity conditions of an area.

[0016] Second, the method of the present invention can effectively improve the interference resistance and accuracy of monitoring results. Specifically, the core of the method lies in the differential processing step (step S5), which performs a linear combination operation on the attenuation change of the first frequency band and the attenuation change of the second frequency band. This step not only effectively suppresses frequency-independent public environmental interference, but also partially suppresses frequency-dependent environmental foreign object interference. This improves the signal-to-noise ratio of the input features (differential attenuation variables) of the subsequent inversion model, enabling stable and accurate acquisition of corresponding soil volumetric moisture content data.

[0017] 2. Other beneficial effects or advantages of the present invention will be described in detail in the specific implementation manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without paying any creative labor.

[0019] in: Figure 1 It is a flowchart of the steps of a method for wireless monitoring of environmental data for smart agriculture provided by an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments.

[0021] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention as claimed, but rather merely represents selected embodiments of the present invention. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without creative effort are intended to fall within the scope of protection of the present invention.

[0022] The technical solution of the present invention is described in detail below with reference to the accompanying drawings.

[0023] like Figure 1 According to a first aspect of the present invention, a method for wireless monitoring of environmental data for smart agriculture is provided, comprising the following steps: Step S1: A LoRa transmitter node and a LoRa receiver node are fixedly deployed on both sides of the soil area to be monitored; Step S2: Control the transmitting node to transmit wireless signals in a first working frequency band and a second working frequency band respectively, wherein the first working frequency band is different from the second working frequency band; Step S3: The receiving node receives and records the first received signal strength and the second received signal strength corresponding to the wireless signals in the first working frequency band and the second working frequency band; Step S4: obtaining a first baseline signal strength and a second baseline signal strength that were previously measured and stored under a soil baseline moisture condition; and calculating a first frequency band attenuation change based on a difference between the first received signal strength and the first baseline signal strength, and calculating a second frequency band attenuation change based on a difference between the second received signal strength and the second baseline signal strength; Step S5: performing a linear combination operation on the attenuation variation of the first frequency band and the attenuation variation of the second frequency band to obtain a differential attenuation variable, so as to significantly suppress frequency-independent public environmental interference and partially suppress frequency-dependent environmental foreign body interference; Step S6: Substituting the differential attenuation variable into a pre-established soil moisture inversion model used to characterize the mapping relationship between the differential attenuation variable and the soil volumetric water content, and calculating the current soil volumetric water content VWC value; Step S7: making a judgment based on the volumetric water content VWC value or its rate of change over time, and generating a warning message when the judgment result exceeds a preset range.

[0024] Through the above technical solution, first, the method of the present invention can achieve non-contact, large-scale soil moisture monitoring. Specifically, the method of the present invention only requires the deployment of LoRa transmitting nodes and LoRa receiving nodes above the soil area to be measured. The soil moisture is inverted by analyzing the transmission characteristics of the wireless signal between the nodes, without the need to deploy physical sensors in the soil. This not only effectively avoids physical damage to the physical sensors caused by agricultural machinery rolling, soil corrosion, or root entanglement, but also significantly reduces hardware costs and deployment and maintenance difficulties. In addition, the monitoring results obtained by a single pair of nodes are the comprehensive effects of the area below their signal propagation path. Compared with the point-type (buried point) measurement results of physical sensors, they are more representative of the overall humidity conditions of an area.

[0025] Second, the method of the present invention can effectively improve the interference resistance and accuracy of monitoring results. Specifically, the core of the method lies in the differential processing step (step S5), which performs a linear combination operation on the attenuation change of the first frequency band and the attenuation change of the second frequency band. This step not only effectively suppresses frequency-independent public environmental interference, but also partially suppresses frequency-dependent environmental foreign object interference. This improves the signal-to-noise ratio of the input features (differential attenuation variables) of the subsequent inversion model, enabling stable and accurate acquisition of corresponding soil volumetric moisture content data.

[0026] It should be noted that in the differential processing step, for common environmental interference that is independent of frequency, such as the impact of atmospheric temperature and humidity changes on the node circuit itself and the slow drift of antenna performance, the impact of these common environmental interference on the signals in the two frequency bands is common-mode (similar), and can be effectively offset by differential operations. For frequency-dependent environmental foreign object interference, such as foreign objects such as stones and plant roots in the soil, these environmental foreign object interferences have different attenuation effects on signals of different frequencies. The linear combination operation can largely separate the impact of these interference factors by processing the changes in the two frequency bands, so that the final differential attenuation variable is mainly sensitive to changes in soil moisture.

[0027] In one embodiment of the present invention, in step S2 of the present invention, the first operating frequency band and the second operating frequency band are selected from licensed frequency bands specified in the LoRa protocol, and the frequency difference between the two frequency bands is greater than 40 MHz.

[0028] In this embodiment, first, the first working frequency band and the second working frequency band are both selected from the licensed frequency bands specified in the LoRa protocol. In this way, it can be ensured that the method of the present invention can avoid interference from other electromagnetic systems as much as possible while complying with relevant laws and regulations.

[0029] Second, setting a larger frequency interval significantly amplifies the differences in soil moisture's effects on signals of different frequencies, thereby optimizing the linear combination operation—most effectively extracting soil moisture information while suppressing environmental interference. This is because when electromagnetic waves propagate through a medium, higher-frequency waves are more sensitive to moisture. Therefore, the difference in attenuation changes between the two frequency bands can be used to invert the corresponding soil moisture.

[0030] The reason why the frequency difference between the two frequency bands is greater than 40 MHz is to ensure that the attenuation of electromagnetic waves in the two frequency bands in response to changes in soil moisture when passing through the soil is large enough and can be accurately measured. After comprehensive consideration and experiments, the lower limit of the frequency difference is set to 40 MHz.

[0031] In one embodiment of the present invention, in step S4 of the present invention, the first baseline signal strength and the second baseline signal strength are obtained by controlling the transmitting node and the receiving node to perform signal measurement when the soil is at a known soil moisture after the node deployment is completed.

[0032] In this implementation, the specific method and timing for obtaining the first and second baseline signal strengths are defined. This allows for precise relative measurements, eliminates systematic errors, and improves absolute accuracy. Specifically, by measuring the baseline at an actual deployment site under conditions of known soil moisture, a highly accurate reference benchmark adapted to the actual deployment conditions is established for all subsequent measurements. This ensures that the resulting inferred soil volumetric moisture content has high absolute accuracy, rather than simply reflecting relative moisture changes.

[0033] In one embodiment of the present invention, in step S5 of the present invention, the linear combination operation specifically includes: subtracting the product of the calibration coefficient and the attenuation change of the second frequency band from the attenuation change of the first frequency band to calculate the differential attenuation variable : Where, is the attenuation change of the first frequency band, is the calibration coefficient, is the attenuation change of the second frequency band.

[0034] In this embodiment, accurate and calculable separation of interference components can be achieved through such linear combination calculation method, wherein, The calculation formula can quantitatively and adjustably separate the target component (change caused by soil moisture) and the interference component from the total attenuation signal, and by setting the coefficients , which can achieve maximum suppression of specific interference.

[0035] In this embodiment, the design principle of the calculation formula of the differential attenuation variable is: assuming that the interference to the two frequency band signals includes common interference (independent of frequency, that is, the impact of the common interference factor on different frequency signals is the same or very similar) and differential interference (for example, moisture in the soil), in the above formula, the common interference is and It appears as a component of similar magnitude (common mode component). During the calculation, if the value of k is appropriate, these common-mode components will cancel each other out due to subtraction. However, the changes caused by soil moisture will not be completely offset due to its special relationship with frequency (reflected by the soil dielectric constant), and will therefore be retained and highlighted.

[0036] In one embodiment of the present invention, the calibration coefficient of the present invention can be determined based on a set physical model, which is used to characterize the functional relationship between the attenuation of the signal by environmental foreign matter and the signal frequency. The calculation expression is: Where, To set the functional relationship of the physical model, is the center frequency of the first operating frequency band, is the center frequency of the second operating frequency band.

[0037] Through this implementation, the coefficient k is transformed from an abstract mathematical parameter that requires "trial and error" to a physical quantity with clear physical meaning and can be determined by theoretical or experimental means. In this way, for different application environments (for example, rocky areas, sandy soils, and crop areas with well-developed root systems), it is only necessary to determine the foreign matter attenuation function corresponding to the environment. , the calibration coefficient most suitable for the environment can also be calculated through the calculation formula of k.

[0038] That is, this embodiment defines a general framework to indicate the solution to the key problem of "how to find the appropriate k value", that is, by studying and characterizing the frequency characteristics of foreign matter attenuation in the environment. In this way, the anti-interference performance of the method of the present invention no longer relies on blind mathematical fitting, but is based on an understanding of the physical mechanism of the interference source. By specifically modeling or measuring the electromagnetic characteristics of major environmental foreign objects (such as common local stone types and crop roots), a coefficient k that can best offset this type of interference is designed, thus ensuring that the differential processing step has optimal interference suppression capabilities for specific environments.

[0039] In this embodiment, it should be noted that the formula The design principle and derivation process can be: The attenuation of radio waves by environmental foreign matter in the soil (stones, roots, debris, etc.) has a certain functional relationship with the frequency f of the radio waves. We define this relationship as the characteristic function This means that for a given foreign body, its frequency is The attenuation of the signal can be expressed as , for frequency The attenuation of the signal can be expressed as , where B is a proportional factor related to the size, density and other properties of the foreign body. In , the attenuation difference caused by foreign matter is included. A suitable k should be selected so that the foreign matter interference term is eliminated after the differential operation. Based on this, if the attenuation caused by foreign matter is completely offset, that is, ; This formula can be transformed into .

[0040] In one embodiment of the present invention, the functional relationship of the present invention is can be modeled as an n-th power function of the signal frequency, that is, The calculation expression is: ; Where n is the power exponent pre-calibrated through experiments.

[0041] In this embodiment, the functional relationship A specific definition is made, clarifying that it is an nth-order power function of the modeled signal frequency. In this way, first, a very simple and operational calculation method for the calibration coefficient k is provided, which simplifies the calculation of k from a process that may require very complex function fitting or measurement to a power operation that only requires two known frequency points and a calibration parameter n. This can achieve a fast, simple and low-cost determination of the k value, thereby enhancing the feasibility and generalizability of the method of the present invention.

[0042] Second, although the power function model Mathematically simple, this formula effectively fits the frequency response of electromagnetic waves interacting with many common foreign objects (those with dimensions smaller than or close to the wavelength). Furthermore, this formula eliminates the proportionality constant that may exist in a power function. We don't need to know the absolute attenuation; only the ratio of the attenuation at two frequencies is required. This ratio is precisely what is needed for the calibration factor k, making the model very simple, dependent only on the frequency ratio and the exponent n. This means that a simple model captures the key contradictions of interference physics, achieving sufficiently high interference suppression with minimal computational complexity, making it ideal for implementation on resource-constrained IoT devices.

[0043] In one embodiment of the present invention, in step S6 of the present invention, the soil moisture inversion model is obtained by training a machine regression algorithm, and the data used to train the machine regression algorithm are the real soil moisture values ​​measured by traditional soil moisture sensors by synchronously collecting multiple sets of differential attenuation variables and their corresponding values ​​after the node is deployed.

[0044] This allows the soil moisture inversion model to be trained locally, combining data from the actual application scenario, regardless of the application scenario (sand, clay, loam, or crop type), to generate an inversion model best suited for the specific environment. In other words, by synchronously collecting local data after deployment, the soil moisture inversion model is calibrated within the real-world application scenario. The trained model inherently incorporates the combined influence of all environmental factors, including soil type, structure, and even the current crop growth status, outputting the most accurate moisture value specific to that particular site. This effectively addresses the pain point of low accuracy in general-purpose models.

[0045] In this implementation, the goal is to establish a supervised learning framework in which the input features are the differential decay variables calculated as above. , the target label is the true soil moisture value measured by the traditional soil moisture sensor, and the goal is to learn a function H such that True VWC, this function H is the soil moisture inversion model.

[0046] In one embodiment of the present invention, the machine learning regression algorithm of the present invention may be a support vector regression algorithm or a gradient boosting decision tree algorithm.

[0047] In this implementation, because deploying traditional sensors in farmland to collect large amounts of real-world moisture data is time-consuming, labor-intensive, and expensive, the dataset used to train the inversion model is likely a small sample size (e.g., only tens to hundreds of data points). The SVR (Support Vector Regression) and GBDT (Gradient Boosting Decision Tree) algorithms selected for this implementation are exemplary of their ability to build robust, high-precision models even with small sample sizes. SVR and GBDT effectively handle the complex, nonlinear relationship between the differential attenuation variable and VWC, accurately capturing this complex mapping and ensuring high accuracy in the inversion model.

[0048] In one embodiment of the present invention, the determination in step S7 of the present invention may specifically include: The current soil volumetric water content (VWC) is compared with a preset drought threshold; if it is lower than the threshold, a drought warning is generated; and / or the rate of change of the current soil volumetric water content (VWC) per unit time is calculated; if the rate of change exceeds a preset water loss rate threshold or water filling rate threshold, a corresponding rapid water loss warning or rapid water filling warning is generated.

[0049] This allows for comprehensive, multi-dimensional early warning assurance based on the current soil volumetric water content (VWC) value and its rate of decline. Drought thresholds provide a static warning, preventing crops from reaching the point of permanent wilting and ensuring basic survival needs. Rapid water loss or rapid water refill warnings are dynamic, identifying sudden anomalies (including rapid water loss and refill) based on VWC changes.

[0050] According to a second aspect of the present invention, there is also provided a wireless monitoring system for environmental data for smart agriculture, which is used to perform the wireless monitoring method for environmental data for smart agriculture according to any one of the technical solutions in the first aspect of the present invention. The wireless monitoring system for environmental data for smart agriculture includes: Multiple pairs of fixedly deployed LoRa node groups, each pair corresponding to a monitoring sub-area, include a transmitting node and a receiving node; wherein the transmitting node is configured to respond to control instructions and transmit wireless signals alternately on a preset first working frequency band and a second working frequency band; the receiving node is configured to receive wireless signals and measure the received signal strength on the first working frequency band and the second working frequency band; A central processing unit is in communication with the receiving node and stores therein the first baseline signal strength, the second baseline signal strength, the calibration coefficient k, and the soil moisture inversion model; the central processing unit is configured to perform the following operations in sequence: receiving a first received signal strength and a second received signal strength from a receiving node; Calculating the attenuation variation of the first frequency band, the attenuation variation of the second frequency band, and the differential attenuation variable; The differential attenuation variable is input into the soil moisture inversion model to obtain the soil volume water content (VWC) value, and it is determined whether the VWC value or its rate of change exceeds the preset range. If so, the early warning module is triggered.

[0051] In this implementation, multiple pairs of fixed LoRa nodes can divide a large area of ​​farmland into several monitoring sub-areas. Each pair of nodes is responsible for measuring a "line" of area. The information from these multiple "lines" is aggregated and integrated by the central processing unit to form humidity situational awareness for the entire farmland. This greatly expands the monitoring capacity and capabilities of a single system, making it suitable for large-scale farmland scenarios in smart agriculture.

[0052] Second, the wireless environmental data monitoring system for smart agriculture of this embodiment can not only realize non-contact, large-scale soil moisture monitoring, but also effectively improve the anti-interference and accuracy of the monitoring results.

[0053] The method of the present invention is further described below with reference to an exemplary embodiment.

[0054] 1. Scenario and system deployment

[0055] 1. Application scenarios A flat farmland planted with corn, with an area of ​​about 500m×1000m.

[0056] 2. Monitoring objectives The average volumetric water content (VWC) of the 0-20cm soil layer of the farmland is inverted and an early warning is issued in case of drought.

[0057] 3. Node deployment: Transmitter node: installed on a fixed pole on one side of the farmland, with the antenna 3.5m above the ground; Receiving node: Installed on a fixed pole on the other side of the farmland, with the antenna 3.5m above the ground; Node spacing: The straight-line distance between two devices (i.e., the path length of the wireless link) is set to 500m; Frequency band selection: Select two frequency points in the CN470 frequency band, among which the first frequency band =470MHz, second frequency band =510MHz.

[0058] Calibration coefficient k calibration: Through preliminary experiments, it was determined that the attenuation of the signal by foreign objects such as stones and roots in the farmland environment is roughly proportional to the square of the frequency (i.e. n=2). Therefore, the calibration coefficient = (470 / 510) 2 ≈0.85.

[0059] 2. Data Processing and Calculation Process The system has previously recorded baseline signal strengths when the soil is dry: the first baseline signal strength = -115dBm, the second baseline signal strength = -118dBm.

[0060] Step 1: Signal Measurement In a certain measurement, the receiving node measured: The first received signal strength in the 470MHz band is -122dBm; The second received signal strength of the 510 MHz frequency band is -128 dBm.

[0061] Step 2: Calculate the attenuation change (-122) - (-115) = -7dB; (-128)-(-118)=-10dB; (Negative values ​​indicate that the signal has attenuated relative to the dry baseline state and the intensity has weakened) Step 3: Differential processing Compute the differential decay variable: (-7)-0.85×(-10)=+1.5dB.

[0062] Step 4: Soil moisture retrieval Assume that a simple linear inversion model has been obtained through machine learning training: VWC=0.5*ΔL_diff+15; Then, the calculated Substituting into the model, we get VWC=15.75%.

[0063] Step 5: Early warning judgment The system's preset drought threshold is 16%; The calculated VWC = 15.9%, which is lower than the preset drought threshold; The system automatically generates a drought warning message, indicating the need to irrigate the area.

[0064] 3. Example and Conclusion 1. Example In this example, the original signal is significantly attenuated in both frequency bands ( =-7dB, = -10dB). If only a single frequency band is considered, it's impossible to determine whether the 3dB of additional attenuation is due to soil wetting or other interference (such as plant growth or the presence of new foreign objects). Using this technical solution's multi-band differential processing, a positive ΔL_diff value (+1.5dB) is obtained. When this value is input into the inversion model, the output soil moisture is 15.9%. This indicates that after eliminating common interference and most foreign object interference, the change in link attenuation is primarily due to an increase in soil dielectric constant, meaning an increase in soil moisture.

[0065] 2. Conclusion First, this example clearly demonstrates how to convert the physical layer signal (RSSI) of a LoRa link into a physically meaningful soil moisture value for agricultural early warning. The entire calculation process is logically rigorous and step-by-step.

[0066] Second, the differential calculation in this example is a key step. It successfully suppresses signal variations insensitive to soil moisture (such as foreign matter and public environmental interference) while retaining and highlighting components sensitive to soil moisture.

[0067] Third, the system ultimately outputs a clear and actionable conclusion, demonstrating its great application potential as a low-cost, large-scale, and probe-free soil moisture monitoring solution.

[0068] The above are only specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions within the technical scope disclosed by the present invention should be covered by the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A wireless monitoring method for environmental data in smart agriculture, characterized in that: The steps include: Step S1: A LoRa transmitter node and a LoRa receiver node are fixedly deployed above both sides of the soil area to be monitored; Step S2: Control the transmitting node to transmit wireless signals in a first working frequency band and a second working frequency band respectively, wherein the first working frequency band is different from the second working frequency band; Step S3: The receiving node receives and records the first received signal strength and the second received signal strength corresponding to the wireless signals in the first working frequency band and the second working frequency band; Step S4: obtaining a first baseline signal strength and a second baseline signal strength that were previously measured and stored under a soil baseline moisture state; and calculating a first frequency band attenuation change based on a difference between the first received signal strength and the first baseline signal strength, and calculating a second frequency band attenuation change based on a difference between the second received signal strength and the second baseline signal strength; Step S5: performing a linear combination operation on the attenuation variation of the first frequency band and the attenuation variation of the second frequency band to obtain a differential attenuation variable, so as to significantly suppress frequency-independent public environmental interference and partially suppress frequency-dependent environmental foreign body interference; Step S6: Substituting the differential attenuation variable into a pre-established soil moisture inversion model for characterizing the mapping relationship between the differential attenuation variable and the soil volumetric water content, and calculating the current soil volumetric water content VWC value; Step S7: making a judgment based on the volumetric water content VWC value or its rate of change over time, and generating a warning message when the judgment result exceeds a preset range.

2. The wireless monitoring method for environmental data of smart agriculture according to claim 1, characterized in that: In step S2, the first operating frequency band and the second operating frequency band are selected from licensed frequency bands specified in the LoRa protocol, and the frequency difference between the two frequency bands is greater than 40 MHz.

3. The wireless monitoring method for environmental data of smart agriculture according to claim 1, characterized in that: In step S4, the first baseline signal strength and the second baseline signal strength are obtained by controlling the transmitting node and the receiving node to perform signal measurement when the soil is at a known soil moisture after the node deployment is completed.

4. The wireless monitoring method for environmental data of smart agriculture according to claim 1, characterized in that: In step S5, the linear combination operation specifically includes: The differential attenuation variable is calculated by subtracting the product of the calibration coefficient and the attenuation change of the second frequency band from the attenuation change of the first frequency band. : Where, is the attenuation change of the first frequency band, is the calibration coefficient, is the attenuation change of the second frequency band.

5. The wireless monitoring method for environmental data of smart agriculture according to claim 4, characterized in that: The calibration coefficient is determined based on a set physical model, which is used to characterize the functional relationship between the attenuation of the signal by environmental foreign matter and the signal frequency. The calculation expression is: Where, To set the functional relationship of the physical model, is the center frequency of the first operating frequency band, is the center frequency of the second operating frequency band.

6. The wireless monitoring method for environmental data of smart agriculture according to claim 5, characterized in that: The functional relationship is modeled as the nth power function of the signal frequency, that is, The calculation expression is: Where n is the power exponent pre-calibrated through experiments.

7. The wireless monitoring method for environmental data of smart agriculture according to claim 1, characterized in that: In step S6, the soil moisture inversion model is obtained by training the machine regression algorithm. The data used to train the machine regression algorithm are the real soil moisture values ​​measured by traditional soil moisture sensors by synchronously collecting multiple groups of differential attenuation variables and their corresponding ones after the node is deployed.

8. The wireless monitoring method for environmental data of smart agriculture according to claim 7, characterized in that: The machine regression algorithm is a support vector regression algorithm or a gradient boosting decision tree algorithm.

9. The wireless monitoring method for environmental data of smart agriculture according to claim 1, characterized in that: The determination in step S7 specifically includes: Comparing the current volumetric water content (VWC) of the soil with a preset drought threshold, and generating a drought warning if the VWC is lower than the threshold; and / or, The rate of change of the current soil volumetric water content (VWC) per unit time is calculated. If the rate of change exceeds the preset water loss rate threshold or water filling rate threshold, a corresponding rapid water loss warning or rapid water filling warning is generated.

10. A wireless monitoring system for environmental data used in smart agriculture, characterized in that: Used to execute the wireless monitoring method for environmental data for smart agriculture according to any one of claims 1 to 9, the wireless monitoring system for environmental data for smart agriculture comprises: Multiple pairs of fixedly deployed LoRa node groups, each pair of node groups corresponding to a monitoring sub-area, including a transmitting node and a receiving node; wherein the transmitting node is configured to respond to control instructions and alternately transmit wireless signals on a preset first working frequency band and a second working frequency band; the receiving node is configured to receive the wireless signals and measure the received signal strength on the first working frequency band and the second working frequency band; A central processing unit is communicatively connected to the receiving node and internally stores the first baseline signal strength, the second baseline signal strength, the calibration coefficient k, and the soil moisture inversion model; the central processing unit is configured to perform the following operations in sequence: receiving a first received signal strength and a second received signal strength from the receiving node; Calculating the attenuation variation of the first frequency band, the attenuation variation of the second frequency band, and the differential attenuation variable; The differential attenuation variable is input into the soil moisture inversion model to obtain the soil volume water content (VWC) value, and it is determined whether the VWC value or its rate of change exceeds a preset range. If so, an early warning module is triggered.

Images