Soil humidity detection method, electronic equipment and computer program product

Abstract

The invention provides a soil humidity detection method, electronic equipment and a computer program product, a radar technology is adopted to replace a traditional contact sensor, radar waves can penetrate through the surface of soil, characteristic parameters are extracted through reflected signals, and the soil humidity is detected by extracting four key parameters including attenuation slope, phase deviation, energy center-of-mass difference and time delay extension. A multi-dimensional feature space is constructed, and the limitation that a single parameter is easily interfered by the environment is overcome. Humidity detection models are respectively established for sandy soil, clay and loam, limitation of a single model is avoided, and high-precision, high-adaptability and low-cost soil humidity detection is realized.

Description

Technical Field

[0001] This application relates to the field of soil moisture detection technology, and more specifically, to a soil moisture detection method, electronic device, and computer program product. Background Technology

[0002] Water is one of the most important components of soil, and its content directly determines the quality of the soil. As the most basic hydrological state variable, soil moisture content plays a crucial role in many fields, including plant growth, agricultural production, soil biochemical reactions, water resource protection, soil erosion control, and land management in arid areas.

[0003] Traditional measurement methods (such as gravimetric moisture method, resistance method, soil moisture meter method, etc.) can ensure measurement accuracy, but they have many limitations: they can damage the observed object, and they also face problems such as difficulty in assimilating data from different types of instruments, limited spatiotemporal resolution, and high cost.

[0004] In the field of microwave remote sensing, various ground-based, airborne, and spaceborne systems, such as synthetic aperture radar and microwave radiometers, as well as active and passive remote sensing systems, have been applied to soil moisture monitoring. However, these detection methods still have technical problems such as low detection accuracy. Summary of the Invention

[0005] The purpose of this application is to provide a soil moisture detection method, electronic device, and computer program product to solve the technical problem of low detection accuracy in existing soil moisture detection schemes.

[0006] This application provides a soil moisture detection method, including: Radar information is obtained by detecting the soil in the target area. Based on radar information, attenuation slope, phase shift, energy centroid difference, and time delay spread are extracted; Soil types are determined based on attenuation slope, phase shift, energy-centroid difference, and time delay spread; soil types include sandy soil, clay soil, and loam. Determine the target moisture detection model based on soil type; The attenuation slope, phase shift, energy-centroid difference, and time delay extension are input into the target humidity detection model to obtain the detection results.

[0007] In the above technical solution, radar technology is used to replace traditional contact sensors. Radar waves can penetrate the soil surface, and characteristic parameters are extracted through reflected signals. By extracting four key parameters—attenuation slope, phase shift, energy-centroid difference, and time delay spread—a multi-dimensional feature space is constructed, overcoming the limitation of single parameters being susceptible to environmental interference. Separate moisture detection models are established for sandy soil, clay, and loam soil, avoiding the limitations of a single model and achieving high-precision, highly adaptable, and low-cost soil moisture detection.

[0008] In some alternative implementations, the attenuation slope is: ; Among them, t i Let be the sampling point at time i, representing the signal propagation time; y i For t i The amplitude of the channel impulse response at time t; For all t within the time window i The mean; For the corresponding time window y i The mean of N; N is the number of sampling points within the time window.

[0009] In the above technical solution, the attenuation slope represents the rate at which the energy of a UWB pulse signal decays over time as it propagates in the soil, and is directly related to the absorption intensity of electromagnetic waves by water. The higher the water content, the faster the electromagnetic waves attenuate, and the larger the absolute value of the slope.

[0010] In some alternative implementations, the phase offset is: ; Where ∠(•) represents the instantaneous phase of the signal; CIR wet The CIR complex signal at the current humidity; CIR dry CIR for pre-measured dry soil.

[0011] In the above technical solution, phase shift occurs because the change in the real part of the dielectric constant of wet soil causes a phase shift in the received signal, reflecting the polarization characteristics of the medium. Increased moisture content increases the dielectric constant, leading to phase lag.

[0012] In some alternative implementations, the centroid difference is: ; in, f k For the k-th frequency point; | Y ( f k | is the signal at a frequency fk The amplitude at that point; K This represents the number of sampling points in the frequency domain.

[0013] In the above technical solution, the energy centroid shift, or the shift in the center of gravity of the signal spectrum energy distribution, reflects the selective absorption of different frequency components by water. Water tends to absorb high-frequency components, causing the energy centroid to shift towards lower frequencies.

[0014] In some alternative implementations, the latency is spread as follows: ; in, τ i For the first i The time delay of the multipath components; h i This represents the complex amplitude of the corresponding multipath component; μ τ This is the average multipath delay. ; M The number of significant multipath paths.

[0015] In the above technical solutions, time delay spread is caused by uneven moisture distribution, which exacerbates the multipath effect and increases the time delay spread value.

[0016] In some alternative implementations, soil type is determined based on attenuation slope, phase shift, energy centroid difference, and time delay spread, including: If the centroid difference is greater than the first set value, then the soil type is sandy soil.

[0017] In the above technical solution, sand has a low dielectric constant and slow attenuation of high-frequency components, causing its centroid to shift towards higher frequencies. Therefore, if the centroid difference is greater than a first set value, for example... f If 3 > 3.8 GHz, then the soil type is sandy soil.

[0018] In some alternative implementations, soil type is determined based on attenuation slope, phase shift, energy centroid difference, and time delay spread, including: If the time delay spread is greater than the second set value, the soil type is clay.

[0019] In the above technical solution, the distribution of clay moisture is uneven, the multipath effect is significant, and the time delay spread is large. Therefore, if the time delay spread is greater than the second set value, for example... f If 4 > 1.7 ns, then the soil type is clay.

[0020] In some alternative implementations, soil type is determined based on attenuation slope, phase shift, energy centroid difference, and time delay spread, including: If the ratio of the attenuation slope to the phase shift is greater than the third set value, then the soil type is loam.

[0021] In the above technical solution, the ratio of attenuation slope (overall absorption) to phase shift (dielectric properties) reflects the soil transition characteristics. Therefore, if the ratio of attenuation slope to phase shift is greater than the third set value, for example... f 1 / f If 2 > 0.6, then the soil type is loam.

[0022] In some alternative implementations, a target moisture detection model is determined based on soil type, including: If the soil type is sandy soil, the target moisture detection model is a polynomial model; If the soil type is clay, the target moisture detection model is a lightweight neural network model; If the soil type is loam, the target moisture detection model is a hybrid model of a multinomial model and a lightweight neural network model.

[0023] In the above technical solution, the UWB radar soil moisture monitoring system employs three different machine learning models to retrieve moisture based on the characteristics of different soil types (sand, clay, and loam).

[0024] For sandy soil, a multinomial model is used for moisture detection. This model is based on a statistical model using multinomial regression, and calculates moisture content by fitting the nonlinear relationship between the feature and moisture. It is suitable for sandy soil with gradually changing dielectric properties. The multinomial model is as follows: ; in, θ v Water content by volume β i represents the fitting coefficient.

[0025] The polynomial model has low computational complexity, requiring only multiplication and addition operations, and can be implemented directly on an MCU without a dedicated accelerator. Since the dielectric constant of sand has an approximately quadratic relationship with moisture content, the polynomial fitting error is <1.5%.

[0026] For clay, a lightweight neural network model is used for moisture detection. This is a miniature neural network designed specifically for clay, with the number of parameters controlled to within 1000 to balance accuracy and power consumption. The lightweight neural network model consists of an input layer, hidden layers, and an output layer. The input layer has 4 neurons (corresponding to a four-dimensional feature vector); the hidden layer has 4 neurons using the ReLU activation function; and the output layer has 1 neuron (linearly activated, directly outputting the moisture value). The lightweight neural network model is suitable for describing the complex relationship between the dielectric constant of clay and moisture (error <1.8%). During model training, a clay-specific dataset (moisture range of 15-40%) needs to be pre-trained.

[0027] For loam, the moisture detection model employs a hybrid architecture combining polynomial and lightweight neural networks, adapting to the transitional characteristics of loam through weighted fusion. Specifically, the polynomial branch fits the linear component dominated by sand, while the neural network branch captures nonlinear features related to clay. The weights can be dynamically adjusted when the proportions of sand and clay in the loam vary. The hybrid model is as follows: ; Where the weight α∈[0.3,0.7].

[0028] An electronic device provided in this application includes a processor and a memory, wherein the memory stores machine-readable instructions executable by the processor, and the machine-readable instructions, when executed by the processor, perform any of the methods described above.

[0029] This application provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of any of the methods described above. Attached Figure Description

[0030] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 A flowchart illustrating the steps of a soil moisture detection method provided in this application embodiment is provided below; Figure 2 This is a schematic diagram of a possible structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0032] The technical solutions in the embodiments of this application will now be described with reference to the accompanying drawings.

[0033] Please refer to Figure 1 , Figure 1 A flowchart of a soil moisture detection method provided in this application embodiment includes: Step S1: Use radar to detect the soil in the target area and obtain radar information; The radar in question is a UWB radar, a wireless sensing technology that utilizes nanosecond-level extremely narrow pulses or wideband continuous waves, typically with a bandwidth ≥500MHz or a relative bandwidth >20%. UWB radar features a wideband signal, operating in the 3.1-10.6GHz range (typical value), and can penetrate non-metallic media (such as soil and walls). Due to its extremely short pulses (1-2ns), UWB radar can achieve centimeter-level spatial resolution. The average power consumption of UWB radar can be controlled in the milliwatt (mW) range, making it suitable for IoT devices. The working principle of UWB radar is as follows: it transmits extremely short pulses (e.g., 1.5ns) or wideband frequency-modulated continuous waves (FMCW); it receives multipath echo signals (Channel Impulse Response, CIR) reflected from the target; it calculates the target distance through time delay analysis and inverts the medium characteristics through amplitude / phase changes.

[0034] Specifically, the radar is installed within a radar detection device, which is inserted into the soil of the target area for detection. The radar detection device has a cylindrical probe structure with a conical bottom for easy insertion into the soil (e.g., vertical insertion to 30cm-50cm). The radar detection device supports 360-degree horizontal echo acquisition.

[0035] Step S2: Based on the radar information, extract the attenuation slope, phase shift, energy centroid difference, and time delay spread; The processing of radar information includes adaptive wavelet denoising, baseline correction, soil stratification, and four-dimensional feature extraction.

[0036] Adaptive wavelet denoising: A signal denoising method based on wavelet transform, which adaptively filters out noise while retaining effective signal components by analyzing the energy distribution of the signal in different frequency bands.

[0037] Baseline correction: Eliminates low-frequency drift in the channel impulse response (CIR) (such as DC offset of the circuit or antenna coupling effect) to bring the signal baseline to zero.

[0038] Soil stratification: By analyzing gradient abrupt change points in the CIR signal, different soil stratification interfaces, including the top and bottom surfaces of soil layers, are identified.

[0039] The four characteristics of attenuation slope, phase shift, energy center-to-mass difference, and time delay spread are the core physical quantities in the UWB radar soil moisture monitoring system, reflecting the changes in the moisture of the medium from different dimensions.

[0040] The attenuation slope represents the rate at which the energy of a UWB pulse signal decays over time as it propagates in the soil, and is directly related to the absorption intensity of electromagnetic waves by water.

[0041] The phase shift is caused by the change in the real part of the dielectric constant of wet soil, which leads to a phase shift in the received signal and reflects the polarization characteristics of the medium.

[0042] The energy centroid difference is the shift in the centroid of the energy distribution of the signal spectrum, reflecting the selective absorption of different frequency components by moisture.

[0043] The time delay extension is: the degree of temporal diffusion of the multipath effect is related to the unevenness of water distribution.

[0044] Step S3: Determine the soil type based on the attenuation slope, phase shift, energy-centroid difference, and time delay spread; the soil type includes sandy soil, clay soil, and loam. Sandy soil: coarse-grained (0.05-2mm in diameter), with large pores and high permeability. It has a low dielectric constant and weak ability to absorb electromagnetic waves. Water easily infiltrates, the surface dries quickly, and the moisture gradient in deeper layers is gentle.

[0045] UWB radar response characteristics of sandy soil: Attenuation slope, relatively small absolute value (typical value: -0.15~-0.05 dB / ns), due to low moisture absorption. Poor centroid, high retention of high-frequency components, centroid frequency > 3.8 GHz. Time delay spread, weak multipath effect (< 1.2 ns), due to uniform structure.

[0046] Clay: Fine particles (diameter <0.002mm), small pores, poor permeability. High dielectric constant, water is tightly bound to the particles. Uneven water distribution, easily forming localized high humidity areas.

[0047] UWB radar response characteristics of clay: Significant phase shift lag (-1.2 to -0.8 rad) due to large variations in the real part of the dielectric constant. Time delay spread and significant multipath effect (>1.7 ns) due to uneven moisture distribution. Attenuation slope: Large absolute value (-0.3 to -0.2 dB / ns), indicating strong moisture absorption.

[0048] Loam: A mixture of sand, silt, and clay, with moderate porosity and a balance between permeability and water retention. Its dielectric constant lies between that of sandy soil and clay. Moisture distribution is relatively uniform, but stratification exists.

[0049] UWB radar response characteristics of soil: Typical attenuation slope / phase offset ratio is 0.6–1.2, reflecting transition characteristics. Poor centroid performance, moderate (3.5–3.8 GHz). Moderate delay spread (1.2–1.7 ns).

[0050] Step S4: Determine the target moisture detection model based on the soil type; Among them, the moisture detection model for sand can use polynomial regression. The dielectric constant of sand has an approximately quadratic nonlinear relationship with moisture. The polynomial model can fit this characteristic with low computational complexity and does not require complex training.

[0051] A neural network model can be used to detect the moisture content of clay. The dielectric constant of clay exhibits a highly nonlinear relationship with moisture content (due to uneven moisture distribution and complex dielectric relaxation), which is difficult to fit using traditional polynomial models. However, neural networks can effectively capture these nonlinear characteristics. The neural network model is obtained through pre-training, with its inputs being the sample's decay slope, phase shift, energy-centroid difference, and time delay spread, and its output being the sample's moisture content.

[0052] A hybrid model for detecting soil moisture can be adopted, which combines multinomial regression and neural network weighted fusion. Soil is a transitional type between sandy soil and clay soil, and its dielectric properties exhibit both linear and nonlinear components. The hybrid model accurately adapts to the complex characteristics of soil by dynamically weighting and fusing the two models.

[0053] Step S5: Input the attenuation slope, phase shift, energy centroid difference, and time delay extension into the target humidity detection model to obtain the detection results.

[0054] In this embodiment, radar technology is used instead of traditional contact sensors. Radar waves can penetrate the soil surface, and feature parameters are extracted through the reflected signals. By extracting four key parameters—attenuation slope, phase shift, energy-centroid difference, and time delay spread—a multi-dimensional feature space is constructed, overcoming the limitation of single parameters being susceptible to environmental interference. Separate moisture detection models are established for sandy soil, clay, and loam soil, avoiding the limitations of a single model and achieving high-precision, highly adaptable, and low-cost soil moisture detection.

[0055] In some alternative implementations, the attenuation slope is: ; Among them, t i Let be the sampling point at time i, representing the signal propagation time; y i For t i The amplitude of the channel impulse response at time t; For all t within the time window i The mean; For the corresponding time window y i The mean of N; N is the number of sampling points within the time window.

[0056] In this embodiment, the attenuation slope represents the rate at which the energy of a UWB pulse signal decays over time as it propagates in the soil, and is directly related to the absorption intensity of electromagnetic waves by water. The higher the water content, the faster the electromagnetic waves attenuate, and the larger the absolute value of the slope.

[0057] In some alternative implementations, the phase offset is: ; Where ∠(•) represents the instantaneous phase of the signal; CIRwet The CIR complex signal at the current humidity; CIR dry CIR for pre-measured dry soil.

[0058] In this embodiment, phase shift occurs because the change in the real part of the dielectric constant of wet soil causes a phase shift in the received signal, reflecting the polarization characteristics of the medium. Increased moisture content increases the dielectric constant, leading to phase hysteresis.

[0059] In some alternative implementations, the centroid difference is: ; in, f k For the k-th frequency point; | Y ( f k | is the signal at a frequency f k The amplitude at that point; K This represents the number of sampling points in the frequency domain.

[0060] In this embodiment, the energy centroid shift, the shift in the center of gravity of the signal spectrum energy distribution, reflects the selective absorption of different frequency components by moisture. Moisture tends to absorb high-frequency components, causing the energy centroid to shift to lower frequencies.

[0061] In some alternative implementations, the latency is spread as follows: ; in, τ i For the first i The time delay of the multipath components; h i This represents the complex amplitude of the corresponding multipath component; μ τ This is the average multipath delay. ; M The significant multipath quantity refers to the number of effective multipath signal paths in the echo signal received by UWB radar that have an energy intensity exceeding a set threshold and contribute to soil moisture detection. The method for determining M is as follows: perform a peak search on the amplitude sequence of the channel impulse response (CIR). Paths satisfying the following conditions are considered effective multipaths: Amplitude condition: The peak value is ≥10% of the direct wave amplitude (threshold adjustable).

[0062] Time condition: Delay difference ≥ 1ns (to avoid double counting of similar paths).

[0063] For sandy soil, M is usually 1-3; for clay, M can reach 4-6.

[0064] In this embodiment, time delay spread is amplified by uneven moisture distribution, which exacerbates the multipath effect and increases the time delay spread value.

[0065] In some alternative implementations, soil type is determined based on attenuation slope, phase shift, energy centroid difference, and time delay spread, including: If the centroid difference is greater than the first set value, then the soil type is sandy soil.

[0066] In this embodiment, the sand has a low dielectric constant and its high-frequency components decay slowly, causing the centroid to shift towards higher frequencies. Therefore, if the centroid difference is greater than a first preset value, for example... f If 3 > 3.8 GHz, then the soil type is sandy soil.

[0067] In some alternative implementations, soil type is determined based on attenuation slope, phase shift, energy centroid difference, and time delay spread, including: If the time delay spread is greater than the second set value, the soil type is clay.

[0068] In this embodiment, the clay moisture distribution is uneven, the multipath effect is significant, and the time delay spread is large. Therefore, if the time delay spread is greater than the second preset value, for example... f If 4 > 1.7 ns, then the soil type is clay.

[0069] In some alternative implementations, soil type is determined based on attenuation slope, phase shift, energy centroid difference, and time delay spread, including: If the ratio of the attenuation slope to the phase shift is greater than the third set value, then the soil type is loam.

[0070] In this embodiment, the ratio of attenuation slope (overall absorption) to phase shift (dielectric properties) reflects the soil transition characteristics. Therefore, if the ratio of attenuation slope to phase shift is greater than a third set value, for example... f 1 / f If 2 > 0.6, then the soil type is loam.

[0071] In some alternative implementations, a target moisture detection model is determined based on soil type, including: If the soil type is sandy soil, the target moisture detection model is a polynomial model; If the soil type is clay, the target moisture detection model is a lightweight neural network model; If the soil type is loam, the target moisture detection model is a hybrid model of a multinomial model and a lightweight neural network model.

[0072] In this embodiment of the application, three different machine learning models were used to retrieve moisture in the UWB radar soil moisture monitoring system, taking into account the characteristics of different soil types (sand, clay, and loam).

[0073] For sandy soil, a multinomial model is used for moisture detection. This model is based on a statistical model using multinomial regression, and calculates moisture content by fitting the nonlinear relationship between the feature and moisture. It is suitable for sandy soil with gradually changing dielectric properties. The multinomial model is as follows: ; in, θ v Water content by volume β i represents the fitting coefficient.

[0074] The polynomial model has low computational complexity, requiring only multiplication and addition operations, and can be implemented directly on an MCU without a dedicated accelerator. Since the dielectric constant of sand has an approximately quadratic relationship with moisture content, the polynomial fitting error is <1.5%.

[0075] For clay, a lightweight neural network model is used for moisture detection. This is a miniature neural network designed specifically for clay, with the number of parameters controlled to within 1000 to balance accuracy and power consumption. The lightweight neural network model consists of an input layer, hidden layers, and an output layer. The input layer has 4 neurons (corresponding to a four-dimensional feature vector); the hidden layer has 4 neurons using the ReLU activation function; and the output layer has 1 neuron (linearly activated, directly outputting the moisture value). The lightweight neural network model is suitable for describing the complex relationship between the dielectric constant of clay and moisture (error <1.8%). During model training, a clay-specific dataset (moisture range of 15-40%) needs to be pre-trained.

[0076] For loam, the moisture detection model employs a hybrid architecture combining polynomial and lightweight neural networks, adapting to the transitional characteristics of loam through weighted fusion. Specifically, the polynomial branch fits the linear component dominated by sand, while the neural network branch captures nonlinear features related to clay. The weights can be dynamically adjusted when the proportions of sand and clay in the loam vary. The hybrid model is as follows: ; The weight α ∈ [0.3, 0.7]. The greater the proportion of sand and the smaller the proportion of clay, the larger the value of α; the smaller the proportion of sand and the greater the proportion of clay, the smaller the value of α.

[0077] This application also provides a soil moisture detection device, including: The radar module is used to detect the soil in the target area using radar and obtain radar information; The feature extraction module is used to extract attenuation slope, phase shift, energy centroid difference, and time delay spread based on radar information. The classification module is used to determine soil type based on attenuation slope, phase shift, energy-centroid difference, and time delay spread; the soil types include sandy soil, clay soil, and loam. The mapping module is used to determine the target humidity detection model based on the soil type. The calculation module is used to input the attenuation slope, phase shift, energy centroid difference, and time delay extension into the target humidity detection model to obtain the detection results.

[0078] In some alternative implementations, the radar module is also used to rotate by a set angle to detect the soil in the next area, so as to obtain the detection results of the soil in the next area.

[0079] Figure 2 This illustration shows a possible structure of an electronic device provided in an embodiment of this application. (Refer to...) Figure 2 The electronic device includes a processor, memory, and a communication interface, which are interconnected and communicate with each other via a communication bus and / or other forms of connection mechanism (not shown).

[0080] The memory includes one or more (only one is shown in the figure), which can be, but is not limited to, Random Access Memory (RAM), Read Only Memory (ROM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), etc. The processor and other possible components can access the memory, reading and / or writing data to it.

[0081] The processor comprises one or more (only one is shown in the figure), which can be an integrated circuit chip with signal processing capabilities. The aforementioned processor can be a general-purpose processor, including a Central Processing Unit (CPU), a Microcontroller Unit (MCU), a Network Processor (NP), or other conventional processors; it can also be a special-purpose processor, including a Neural-network Processing Unit (NPU), a Graphics Processing Unit (GPU), a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. Furthermore, when there are multiple processors, some can be general-purpose processors, and others can be special-purpose processors.

[0082] The communication interface includes one or more (only one is shown in the figure), which can be used to communicate directly or indirectly with other devices to exchange data. The communication interface may include interfaces for wired and / or wireless communication.

[0083] One or more computer program instructions may be stored in the memory, and the processor may read and execute these computer program instructions to implement the methods provided in the embodiments of this application.

[0084] Understandable. Figure 2 The structure shown is for illustrative purposes only; the electronic device may also include structures that are more complex than those shown. Figure 2 The more or fewer components shown, or having the same Figure 2 The different structures shown. Figure 2 The components shown can be implemented using hardware, software, or a combination thereof. Electronic devices may be physical devices, such as PCs, laptops, tablets, mobile phones, servers, embedded devices, etc., or they may be virtual devices, such as virtual machines, virtualized containers, etc. Furthermore, electronic devices are not limited to a single device; they can also be a combination of multiple devices or a cluster of a large number of devices.

[0085] This application provides a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of any of the methods described above.

[0086] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0087] Furthermore, the units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0088] Furthermore, the functional modules in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0089] In this document, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, without necessarily requiring or implying any such actual relationship or order between these entities or operations.

[0090] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for detecting soil moisture, characterized in that, include: Radar information is obtained by detecting the soil in the target area. Based on the radar information, the attenuation slope, phase shift, energy centroid difference, and time delay spread are extracted. Soil type is determined based on the attenuation slope, phase shift, energy center-of-mass difference, and time delay spread; wherein, the soil type includes sandy soil, clay soil, and loam. Based on the soil type, determine the target humidity detection model; The attenuation slope, phase shift, energy centroid difference, and time delay spread are input into the target humidity detection model to obtain the detection results.

2. The method as described in claim 1, characterized in that, The attenuation slope is: ； Among them, t i Let be the sampling point at time i, representing the signal propagation time; y i For t i The amplitude of the channel impulse response at time t; For all t within the time window i The mean; For the corresponding time window y i The mean of N; N is the number of sampling points within the time window.

3. The method as described in claim 1, characterized in that, The phase shift is: ； Where ∠(•) represents the instantaneous phase of the signal; CIR wet The CIR complex signal at the current humidity; CIR dry CIR for pre-measured dry soil.

4. The method as described in claim 1, characterized in that, The energy centroid difference is: ； in, f k For the k-th frequency point; | Y ( f k | is the signal at a frequency f k The amplitude at that point; K This represents the number of sampling points in the frequency domain.

5. The method as described in claim 1, characterized in that, The delay spread is as follows: ； in, τ i For the first i The time delay of the multipath components; h i This represents the complex amplitude of the corresponding multipath component; μ τ This is the average multipath delay. ; M The number of significant multipath paths.

6. The method as described in claim 1, characterized in that, The determination of soil type based on the attenuation slope, phase shift, energy-centroid difference, and time delay spread includes: If the energy centroid difference is greater than a first set value, then the soil type is sandy soil.

7. The method as described in claim 1, characterized in that, The determination of soil type based on the attenuation slope, phase shift, energy-centroid difference, and time delay spread includes: If the time delay spread is greater than the second set value, then the soil type is clay.

8. The method as described in claim 1, characterized in that, The determination of soil type based on the attenuation slope, phase shift, energy-centroid difference, and time delay spread includes: If the ratio of the attenuation slope to the phase offset is greater than a third set value, then the soil type is loam.

9. The method as described in claim 1, characterized in that, The step of determining the target humidity detection model based on the soil type includes: If the soil type is sandy soil, the target humidity detection model is a polynomial model; If the soil type is clay, the target moisture detection model is a lightweight neural network model; If the soil type is loam, the target humidity detection model is a hybrid model of a multinomial model and a lightweight neural network model.

10. An electronic device, characterized in that, include: A processor and a memory, the memory storing machine-readable instructions executable by the processor, which, when executed by the processor, perform the method as described in any one of claims 1-9.

11. A computer program product, comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method according to any one of claims 1-9.

Images