Dielectric frequency spectrum-based multi-depth greenhouse matrix soil moisture sensing system and method

Abstract

The invention belongs to the technical field of agricultural environment monitoring, and more specifically relates to a multi-depth greenhouse matrix soil moisture content sensing system based on a dielectric frequency spectrum. The system comprises a data acquisition device, at least one matrix sensor and a communication module, the data acquisition device comprises core circuits such as a core controller STM32L431RCT6 and the like, and is used for acquiring data of the matrix sensor; the matrix sensor is in communication connection with the data acquisition device through a 485 serial port and a Modbus RTU slave station protocol, and is used for acquiring the volumetric water content and the conductivity of a matrix; and the communication module adopts a 4G Cat1 module, is in communication connection with the data acquisition device by adopting a UART (Universal Asynchronous Receiver / Transmitter), and interacts with an external Internet of Things cloud platform through an MQTT protocol. The invention provides a system capable of adapting to a complex matrix environment of a greenhouse, realizing multi-depth in-situ high-precision measurement and stably operating soil moisture content monitoring for a long time.

Description

Technical Field

[0001] This invention belongs to the technical field of agricultural environmental monitoring, and more specifically, relates to a multi-depth greenhouse substrate moisture sensing system and method based on dielectric spectrum. Background Technology

[0002] Greenhouse cultivation is an important form of modern agricultural production. Crop growth in greenhouses is highly dependent on the stability of the root environment. Substrate moisture, particularly volumetric water content and electrical conductivity (which characterize nutrient concentration), are key environmental factors affecting root vigor, water absorption capacity, and nutrient uptake. However, greenhouse substrates are typically composed of a complex mixture of soil, organic matter, vermiculite, and perlite. Their properties change dynamically during irrigation, evaporation, crop absorption, and organic matter decomposition, leading to unstable dielectric properties. Currently, most commercially available substrate moisture sensors are based on resistance or single-frequency capacitance principles.

[0003] These types of sensors exhibit significant drift in their measurements under conditions of high ion concentration (high conductivity), changes in organic matter content, and moisture redistribution, making accuracy difficult to guarantee. Therefore, they cannot meet the urgent need for long-term, continuous, in-situ, high-precision monitoring of the substrate environment in greenhouse production. Furthermore, traditional sensors mostly perform single-point measurements, failing to reflect the vertical distribution of soil moisture in the root zone, and lack reliable communication and power supply solutions for long-term continuous monitoring of the greenhouse environment. Summary of the Invention

[0004] The present invention aims to overcome at least one of the defects of the prior art and provide a multi-depth greenhouse substrate moisture sensing system based on dielectric spectrum, which can adapt to the complex substrate environment of greenhouse, realize multi-depth in-situ high-precision measurement, and can operate stably for a long time for moisture monitoring. The present invention also discloses a multi-depth greenhouse substrate moisture perception method based on dielectric spectrum.

[0005] The detailed technical solution of this invention is as follows: A multi-depth greenhouse substrate moisture sensing system based on dielectric spectrum, the system comprising: a data acquisition device, at least one substrate sensor, and a communication module; The data acquisition device includes a core controller STM32L431RCT6, which is used to acquire data from the matrix sensor. The matrix sensor and data acquisition device are connected via a 485 serial port and Modbus RTU slave protocol to collect the volumetric water content and conductivity of the matrix. The communication module adopts a 4G Cat1 module, which is connected to the data acquisition device via UART communication and interacts with the external IoT cloud platform through the MQTT protocol.

[0006] According to a preferred embodiment of the present invention, the matrix sensor comprises a flexible PCB microstrip line probe, a dielectric spectrum measurement circuit, and a core controller STM32F103C8T6, wherein: The dielectric spectrum measurement circuit is electrically connected to the core controller STM32F103C8T6 and the probe. The dielectric spectrum measurement circuit includes a DDS signal generation unit and a signal detection unit. The DDS signal generating unit is used to inject a swept-frequency sinusoidal excitation signal with a frequency range of 10MHz to 80MHz into one electrode of the probe; The signal detection unit is used to receive the response signal propagated through the matrix via another electrode of the probe, and to measure the phase difference Δφ and amplitude ratio A at multiple preset characteristic frequency points. _ratio The signal detection unit is also used to accurately measure the current flowing through the matrix and the voltage across the matrix using an automatic balancing bridge circuit. The flexible PCB microstrip probe can achieve in-situ measurements at four depths of 10cm, 20cm, 30cm, and 40cm without severely damaging the substrate structure, and obtain the phase difference and amplitude ratio. The core controller STM32F103C8T6 has a built-in dielectric spectrum prediction model, which calculates the volumetric water content and conductivity based on the phase difference and amplitude ratio.

[0007] According to a preferred embodiment of the present invention, the flexible PCB microstrip probe includes electrode A and electrode B; Electrode A is a multifunctional measuring electrode located on the far left. Its end is equipped with a dielectric spectrum receiving module, which is used to receive the transmission response of high-frequency excitation signals in the matrix and invert dielectric properties and related physical parameters. The dielectric spectrum receiving module includes a lock-in amplifier, which is used to demodulate the phase difference and amplitude ratio. Electrode B is a dielectric spectrum emitting electrode located at the far right end. A frequency sweeping high-frequency signal excitation module is provided at the bottom, forming a non-common-point capacitive coupling path with electrode A, which is used to inject electromagnetic waves into the matrix to trigger the dielectric response.

[0008] According to a preferred embodiment of the present invention, the data acquisition device is powered by a solar panel and a lithium battery, and features a low-power design and low-voltage protection.

[0009] According to a preferred embodiment of the present invention, the data acquisition device further includes a built-in high-capacity FLASH memory.

[0010] According to a preferred embodiment of the present invention, the dielectric spectrum prediction model incorporates a large amount of laboratory calibration data and a feature screening algorithm to achieve the calculation of volumetric water content and conductivity from the original signal.

[0011] This invention also provides a multi-depth greenhouse substrate moisture sensing method based on dielectric spectrum, using the multi-depth greenhouse substrate moisture sensing system based on dielectric spectrum as described above, characterized in that the method includes: S1. Hardware Deployment: Vertically insert the depth probe of the substrate sensor into the substrate of the cultivation trough, ensuring that the depth measurement point on the probe is located in the main root activity layer of the tomato; fix the data acquisition device to the inner wall of the greenhouse, connect the solar panel, and connect one or more substrate sensors to the data acquisition device via RS485 bus. S2. Trigger Acquisition: When the acquisition time is reached, the data acquisition device, acting as a Modbus master station, sends a data request frame to the designated matrix sensor. S3. Signal Measurement: The core controller of the matrix sensor drives the microstrip probe. The DDS signal generation unit injects a tunable sinusoidal signal into electrode B at a pre-selected characteristic frequency point. The sweep frequency high-frequency signal excitation module of electrode B injects electromagnetic waves into the matrix to trigger the dielectric response. After the signal propagates through the matrix, its phase shifts due to changes in the water state, and its amplitude attenuates due to the influence of solute ions and conductivity. The signal is finally received by the dielectric spectrum receiving module at the end of electrode A, and the phase difference Δφ and amplitude ratio A_ratio of the signal at each frequency point are demodulated by the lock-in amplifier. S4. Parameter calculation: The controller inputs the phase difference and amplitude ratio of the measured characteristic frequency points into the preset dielectric spectrum prediction model to calculate the volumetric water content and conductivity values ​​at four depths in real time. S5. Data Response and Upload: The matrix sensor encapsulates the calculation results in a Modbus RTU response frame and returns them to the data acquisition device. After the core controller of the data acquisition device parses the data, it immediately uploads it to the IoT cloud platform through the communication module and MQTT protocol. If the wireless network signal is poor at this time, the data will be automatically stored in the FLASH memory and the breakpoint will be recorded. The data acquisition device will continue to try to reconnect. After the network is restored, it will automatically re-transmit the historical data from the breakpoint. S6. Data Application: Users can view volumetric water content and conductivity information in real time on the IoT cloud platform, make precise irrigation and fertilization decisions, and achieve water conservation, fertilizer reduction, and increased production and efficiency.

[0012] According to a preferred embodiment of the present invention, in step S4, the real-time calculation of the volumetric water content and conductivity values ​​at four depths specifically refers to: The volumetric water content was obtained by fitting the voltage values ​​flowing through both ends of the matrix, which were measured by an automatic balancing bridge circuit. The conductivity refers to the mathematical relationship between the phase difference and amplitude ratio and the material's dielectric constant, established based on transmission line theory or electromagnetic field propagation model. The space filled with the material is regarded as a transmission line, and its characteristic impedance and propagation constant are closely related to the material's complex dielectric constant. The propagation constant γ is a key bridge, and its imaginary part is the phase constant. Related to phase difference: Δφ= *d, where d is the material thickness; its real part is the attenuation constant. Related to amplitude decay: A _ratio =EXP(- *d), where d is the material thickness; The measured phase difference and amplitude ratio are converted into attenuation constants. and phase constant Then, construct information about the complex permittivity. The system of equations is solved by numerical iteration or a calibration-based model to obtain the real part of the complex permittivity. and the virtual part The imaginary part of the dielectric constant That is, the material's power consumption capacity and its electrical conductivity. Related, that is, = ,in It is the angular frequency of the magnetic field. It is the dielectric constant in a vacuum; The calibration method uses standard samples with known dielectric constants for calibration, and establishes a quantitative relationship model between phase difference and amplitude ratio and dielectric constant, which is used to measure unknown sample matrices.

[0013] According to a preferred embodiment of the present invention, the matrix sensor adopts a three-point or ten-point calibration mechanism, and the corresponding calibration mechanism can be set at the factory via the Modbus protocol, selecting three-point or ten-point calibration.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention has high measurement accuracy and strong anti-interference ability: the core lies in the adoption of the multi-frequency dielectric spectrum measurement principle and the combination of a dedicated dielectric spectrum prediction model. By analyzing the phase and amplitude information of multiple characteristic frequency points, the combined effects of moisture, conductivity and organic matter on the measurement signal can be effectively decoupled, thereby fundamentally overcoming the significant measurement drift problem of traditional single-frequency capacitive or resistive sensors in high-salt and organic matter changing environments, and realizing the accurate perception of key parameters in complex matrix environments.

[0015] (2) The present invention realizes in-situ and multi-depth profile monitoring with rich information: An innovative microstrip line depth probe based on flexible PCB is designed. A single probe can achieve synchronous measurement at different depths. This structure can be deployed in-situ without seriously damaging the matrix structure, and can obtain the complete soil moisture profile information in the crop root area at one time, providing unprecedented data support for accurate irrigation decision-making and overcoming the limitations of traditional single-point measurement.

[0016] (3) The system of the present invention has high reliability and is suitable for long-term deployment: The system-level design fully considers the actual application scenarios in greenhouses. The sensor probe has a sturdy structure; the data interaction adopts the RS485 / Modbus RTU industrial bus protocol with strong anti-interference ability; the data collector has functions such as solar power supply, large-capacity offline caching, and breakpoint resumption, ensuring the continuous operation ability under bad weather and unstable network conditions, and meeting the rigid requirements of long-term and continuous monitoring in greenhouse production.

[0017] (4) The present invention has high expandability and intelligence: The system leaves room for subsequent function expansion. The accurately measured conductivity value can be directly used to infer the dynamic concentration of nutrient ions such as nitrogen, phosphorus, and potassium in the matrix, laying a foundation for precise fertilization. At the same time, the "sensing end" can set the corresponding calibration mechanism through the modbus protocol during factory production, and select three-point or ten-point calibration to improve the accuracy of sensor calibration. On the one hand, it reduces the computing pressure on the cloud, and on the other hand, it enhances the on-site calibration ability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is the structural diagram of the multi-depth greenhouse substrate soil moisture sensing system based on dielectric spectroscopy according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] The following further describes the present disclosure in conjunction with the drawings and embodiments.

[0020] Embodiment 1 Refer Figure 1 to this, this embodiment provides a multi-depth greenhouse substrate soil moisture sensing system based on dielectric spectroscopy, including a data acquisition device, multiple substrate sensors, and a communication module; The product model of the data acquisition device is named SK3431, its core controller is STM32L431RCT6, which is powered by a solar panel and a lithium battery, and has a low-power design and low-voltage protection to ensure continuous operation for more than 20 days in rainy days; it real-time collects the power supply voltage information, and the core controller sets the voltage threshold. When the voltage is lower than this threshold, the single-chip microcomputer is set to enter the low-power mode, the power supply of the communication circuit is turned off, and the data is stored to prevent data loss; The data acquisition device also includes built-in large-capacity FLASH storage, with local caching and breakpoint resume functions to ensure no data loss. The actual storage chip used is an off-chip FLASH (W25Q512), capable of storing 64MB of data, with an SPI interface. When designing the driver, three ordinary I / O pins of the controller can be used to simulate SPI communication. It operates on a 3.3V power supply, with a storage area of ​​0-64MB, and supports byte addressing.

[0021] The data acquisition device is used to collect matrix sensor data from multiple points and depths, and is designed to form a network based on the characteristics of the sensors. It has the characteristics of strong anti-interference ability and long data transmission distance.

[0022] The communication module adopts a 4G Cat1 module, communicates with the data acquisition device via UART, and interacts with the IoT cloud platform via the MQTT protocol to achieve low-power long-distance data transmission. The matrix sensor product model is named SK3103, which includes a flexible PCB microstrip probe, a dielectric spectrum measurement circuit and a core controller STM32F103C8T6. The matrix sensor is used to collect data such as the volumetric water content and conductivity of the matrix, and communicates with the data acquisition device through a 485 serial port and Modbus RTU slave protocol. The dielectric spectrum measurement circuit is electrically connected to the core controller STM32F103C8T6 and the probe. The dielectric spectrum measurement circuit includes a DDS signal generation unit and a signal detection unit. The DDS signal generation unit injects a swept-frequency sinusoidal excitation signal with a frequency range of 10 MHz to 80 MHz into one electrode of the probe. The signal detection unit receives the response signal propagated through the matrix via the other electrode of the probe and measures the phase difference Δφ and amplitude ratio A at multiple preset characteristic frequency points. _ratio The signal detection unit is also used to accurately measure the current flowing through the matrix and the voltage across the matrix using an automatic balancing bridge circuit. The flexible PCB microstrip probe has a compact structure and can achieve in-situ measurements at four depths of 10cm, 20cm, 30cm and 40cm without seriously damaging the matrix structure. It can accurately reflect changes in soil moisture profile and can achieve in-situ continuous monitoring without damaging the matrix structure. The probe includes electrode A and electrode B. Electrode A is a multifunctional measuring electrode located on the far left. Its end is equipped with a dielectric spectrum receiving module, which is used to receive the transmission response of high-frequency excitation signals in the matrix and invert dielectric properties and related physical parameters. The dielectric spectrum receiving module includes a lock-in amplifier, which is used to demodulate the phase difference and amplitude ratio. Electrode B is a dielectric spectrum emission electrode located at the far right end. A sweep frequency high-frequency signal excitation module is provided at the bottom, which forms a non-common point capacitive coupling path with electrode A, and is used to inject electromagnetic waves into the matrix to trigger the dielectric response. The core controller STM32F103C8T6 has a built-in dielectric spectrum prediction model, which calculates the volume water content and conductivity based on the phase difference and amplitude ratio. The dielectric spectrum prediction model incorporates a large amount of laboratory calibration data (pure water, alcohol, KCl solution, tomato matrix) and feature screening algorithms (such as principal component analysis) to calculate the volume water content and conductivity from the original signal.

[0023] Example 2 This embodiment provides a multi-depth greenhouse substrate moisture perception method based on dielectric spectrum, the method comprising: S1. Hardware Deployment: Vertically insert the depth probe of the SK3103 substrate sensor into the substrate of the cultivation trough, ensuring that the 10cm, 20cm, 30cm, and 40cm depth measurement points on the probe are located in the active layer of the tomato's main root system. Fix the SK3431 data acquisition device to the inner wall of the greenhouse, connect the solar panel, and connect one or more SK3103 substrate sensors to the data acquisition device via RS485 bus. After the system is powered on, the SK3431 automatically scans the sensors on the bus (Modbus slave addresses are preset). Users can set the data acquisition interval (e.g., once every 30 minutes) through the cloud platform or local configuration tools.

[0024] S2. Trigger Acquisition: When the acquisition time is reached, the SK3431 data acquisition device, as a Modbus master station, sends a data request frame to the designated SK3103 matrix sensor. S3. Signal Measurement: The first controller (STM32F103) of the SK3103 matrix sensor drives the microstrip probe. The DDS signal generation unit injects a tunable sinusoidal signal into electrode B at pre-selected characteristic frequency points (e.g., 3-5 key frequencies screened by principal component analysis). The frequency sweeping high-frequency signal excitation module of electrode B injects electromagnetic waves into the matrix to trigger the dielectric response. After the signal propagates through the matrix, its phase shifts due to changes in the water state, and its amplitude attenuates due to the influence of solute ions and conductivity. The signal is finally received by the dielectric spectrum receiving module at the end of electrode A, and the phase difference Δφ and amplitude ratio A_ratio of the signal at each frequency point are demodulated by the lock-in amplifier. S4. Parameter Calculation: The controller inputs the phase difference and amplitude ratio of the measured characteristic frequency points into the preset dielectric spectrum prediction model to calculate the volumetric water content and conductivity values ​​at four depths in real time. Specifically, this means: The volumetric water content was obtained by fitting the voltage values ​​flowing through both ends of the matrix, which were measured by an automatic balancing bridge circuit. The conductivity refers to the mathematical relationship between the phase difference and amplitude ratio and the material's dielectric constant, established based on transmission line theory or electromagnetic field propagation model. The space filled with the material is regarded as a transmission line, and its characteristic impedance and propagation constant are closely related to the material's complex dielectric constant. The propagation constant γ is a key bridge, and its imaginary part is the phase constant. Related to phase difference: Δφ= *d, where d is the material thickness; its real part is the attenuation constant. Related to amplitude decay: A _ratio =EXP(- *d), where d is the material thickness; The measured phase difference and amplitude ratio are converted into attenuation constants. and phase constant Then, construct information about the complex permittivity. The system of equations is solved by numerical iteration or a calibration-based model to obtain the real part of the complex permittivity. and the virtual part The imaginary part of the dielectric constant. That is, the material's power consumption capacity and its electrical conductivity. Related, that is, = ,in It is the angular frequency of the magnetic field. It is the dielectric constant in a vacuum; The calibration-based approach can use standard samples with known dielectric constants for calibration, and establish a quantitative relationship model between phase difference and amplitude ratio and dielectric constant, which can be used to measure unknown sample matrices.

[0025] S5. Data Response and Upload: The SK3103 encapsulates the calculation results in a Modbus RTU response frame and returns it to the SK3431. The SK3431's second controller (STM32L431) parses the data and immediately uploads it to the cloud platform via a 4G Cat1 module and the MQTT protocol. If the wireless network signal is weak at this time, the data will be automatically stored in a 64MB FLASH memory, and the breakpoint will be recorded. The collector will continuously attempt to reconnect, and automatically resume historical data transmission from the breakpoint once the network is restored. S6. Data Application: Users can view soil moisture data and historical curves at various depths in real time on the cloud platform, and set alarm rules. Based on accurate profile moisture and conductivity information, precise irrigation and fertilization decisions can be made to achieve water conservation, fertilizer reduction, and increased yield and efficiency.

[0026] The matrix sensor employs a three-point or ten-point calibration mechanism, both of which are built into the program. At the factory, the appropriate calibration mechanism can be set via the Modbus protocol, selecting either three-point or ten-point calibration to improve the accuracy of sensor calibration, as detailed below: After the AD acquisition data is processed, it is filtered, converted to a 0-100 scale, and calibrated according to the hardware specifications. The standard data values ​​of the four matrix sensors in different matrix standard solutions are recorded in Table 1 below: Table 1 Comparison of calibration data for the test matrix sensor

[0027] Based on standard test data, matrix sensors exhibit significant individual differences in hardware, making it impossible to uniformly calibrate their data. Each sensor requires individual production calibration before leaving the factory, during which calibration data is written into the chip's Flash memory. If users require recalibration, they can perform recalibration via serial communication to update the calibration values ​​and improve the accuracy of humidity data collection.

[0028] The data acquisition device communicates with the matrix sensor via a 485 serial port using the Modbus RTU protocol. The sensor address (ModbusAddress) can be reconfigured, and the address setting takes effect upon power-up of the sensor device.

[0029] Calibration can be performed in two modes: three-point calibration (MoiCalMode=0x0000) and ten-point calibration (MoiCalMode=0x0001).

[0030] Table 2 Modbus RTU Protocol for Calibration Modes

[0031] The data acquisition device is configured with the serial port command transmission protocol as shown in Table 2 (hex transmission). The matrix sensor is factory defaulted to three-point calibration. After selecting the calibration mode, calibration values ​​can be sent according to the calibration mode. The final sensor calibration value calculation process is shown below:

[0032] in, This represents the final calibration value of the matrix sensor. This represents the sensor test values ​​collected when the substrate is air. This represents the sensor test value collected from a solution with a matrix water content of 25%. This represents the sensor test values ​​collected when the substrate is water. The matrix volumetric water content is 3%. The matrix volume water content is 25%. X represents the volumetric water content of the substrate, which is 97%, and X is the actual measured value of the test substrate.

[0033] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the technical solutions of the present invention, and are not intended to limit the specific implementation of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the claims of the present invention should be included within the protection scope of the claims of the present invention.

Claims

1. A multi-depth greenhouse substrate moisture sensing system based on dielectric spectrum, characterized in that, The system includes: a data acquisition device, at least one matrix sensor, and a communication module; The data acquisition device includes a core controller STM32L431RCT6, which is used to acquire data from the matrix sensor. The matrix sensor and data acquisition device are connected via a 485 serial port and Modbus RTU slave protocol to collect the volumetric water content and conductivity of the matrix. The communication module adopts a 4G Cat1 module, which is connected to the data acquisition device via UART communication and interacts with the external IoT cloud platform through the MQTT protocol.

2. The multi-depth greenhouse substrate moisture sensing system based on dielectric spectrum according to claim 1, characterized in that, The matrix sensor includes a flexible PCB microstrip probe, a dielectric spectrum measurement circuit, and a core controller STM32F103C8T6, wherein: The dielectric spectrum measurement circuit is electrically connected to the core controller STM32F103C8T6 and the probe. The dielectric spectrum measurement circuit includes a DDS signal generation unit and a signal detection unit. The DDS signal generating unit is used to inject a swept-frequency sinusoidal excitation signal with a frequency range of 10MHz to 80MHz into one electrode of the probe; The signal detection unit is used to receive the response signal propagated through the matrix via another electrode of the probe, and to measure the phase difference Δφ and amplitude ratio A at multiple preset characteristic frequency points. _ratio The signal detection unit is also used to accurately measure the current flowing through the matrix and the voltage across the matrix using an automatic balancing bridge circuit. The flexible PCB microstrip probe can achieve in-situ measurements at four depths of 10cm, 20cm, 30cm, and 40cm without severely damaging the substrate structure, and obtain the phase difference and amplitude ratio. The core controller STM32F103C8T6 has a built-in dielectric spectrum prediction model, which calculates the volumetric water content and conductivity based on the phase difference and amplitude ratio.

3. The multi-depth greenhouse substrate moisture sensing system based on dielectric spectrum according to claim 2, characterized in that, The flexible PCB microstrip probe includes electrode A and electrode B; Electrode A is a multifunctional measuring electrode located on the far left. Its end is equipped with a dielectric spectrum receiving module, which is used to receive the transmission response of high-frequency excitation signals in the matrix and invert dielectric properties and related physical parameters. The dielectric spectrum receiving module includes a lock-in amplifier, which is used to demodulate the phase difference and amplitude ratio. Electrode B is a dielectric spectrum emitting electrode located at the far right end. A frequency sweeping high-frequency signal excitation module is provided at the bottom, forming a non-common-point capacitive coupling path with electrode A, which is used to inject electromagnetic waves into the matrix to trigger the dielectric response.

4. The multi-depth greenhouse substrate moisture sensing system based on dielectric spectrum according to claim 1, characterized in that, The data acquisition device is powered by a solar panel and a lithium battery, and features a low-power design and low-voltage protection.

5. The multi-depth greenhouse substrate moisture sensing system based on dielectric spectrum according to claim 1, characterized in that, The data acquisition device also includes a built-in high-capacity FLASH memory.

6. The multi-depth greenhouse substrate moisture sensing system based on dielectric spectrum according to claim 2, characterized in that, The dielectric spectrum prediction model incorporates a large amount of laboratory calibration data and feature screening algorithms to calculate volumetric water content and conductivity from the original signal.

7. A multi-depth greenhouse substrate moisture sensing method based on dielectric spectrum, using the multi-depth greenhouse substrate moisture sensing system based on dielectric spectrum as described in any one of claims 1 to 6 above, characterized in that, The method includes: S1. Hardware Deployment: Vertically insert the depth probe of the substrate sensor into the substrate of the cultivation trough, ensuring that the depth measurement point on the probe is located in the main root activity layer of the tomato; fix the data acquisition device to the inner wall of the greenhouse, connect the solar panel, and connect one or more substrate sensors to the data acquisition device via RS485 bus. S2. Trigger Acquisition: When the acquisition time is reached, the data acquisition device, acting as a Modbus master station, sends a data request frame to the designated matrix sensor. S3. Signal Measurement: The core controller of the matrix sensor drives the microstrip probe. The DDS signal generation unit injects a tunable sinusoidal signal into electrode B at a pre-selected characteristic frequency point. The sweep frequency high-frequency signal excitation module of electrode B injects electromagnetic waves into the matrix to trigger the dielectric response. After the signal propagates through the matrix, its phase shifts due to changes in the water state, and its amplitude attenuates due to the influence of solute ions and conductivity. The signal is finally received by the dielectric spectrum receiving module at the end of electrode A, and the phase difference Δφ and amplitude ratio A_ratio of the signal at each frequency point are demodulated by the lock-in amplifier. S4. Parameter calculation: The controller inputs the phase difference and amplitude ratio of the measured characteristic frequency points into the preset dielectric spectrum prediction model to calculate the volumetric water content and conductivity values ​​at four depths in real time. S5. Data Response and Upload: The matrix sensor encapsulates the calculation results in a Modbus RTU response frame and returns them to the data acquisition device. After the core controller of the data acquisition device parses the data, it immediately uploads it to the IoT cloud platform through the communication module and MQTT protocol. If the wireless network signal is poor at this time, the data will be automatically stored in the FLASH memory and the breakpoint will be recorded. The data acquisition device will continue to try to reconnect. After the network is restored, it will automatically re-transmit the historical data from the breakpoint. S6. Data Application: Users can view volumetric water content and conductivity information in real time on the IoT cloud platform, make precise irrigation and fertilization decisions, and achieve water conservation, fertilizer reduction, and increased production and efficiency.

8. The method for sensing multi-depth greenhouse substrate moisture based on dielectric spectrum according to claim 7, characterized in that, In step S4, the real-time calculation of the volumetric water content and conductivity values ​​at four depths specifically refers to: The volumetric water content was obtained by fitting the voltage values ​​flowing through both ends of the matrix, which were measured by an automatic balancing bridge circuit. The conductivity refers to the mathematical relationship between the phase difference and amplitude ratio and the material's dielectric constant, established based on transmission line theory or electromagnetic field propagation model. The space filled with the material is regarded as a transmission line, and its characteristic impedance and propagation constant are closely related to the material's complex dielectric constant. The propagation constant γ is a key bridge, and its imaginary part is the phase constant. Related to phase difference: Δφ= *d, where d is the material thickness; its real part is the attenuation constant. Related to amplitude decay: A _ratio =EXP(- *d), where d is the material thickness; The measured phase difference and amplitude ratio are converted into attenuation constants. and phase constant Then, construct information about the complex permittivity. The system of equations is solved by numerical iteration or a calibration-based model to obtain the real part of the complex permittivity. and the virtual part The imaginary part of the dielectric constant That is, the material's power consumption capacity and its electrical conductivity. Related, that is, = ,in It is the angular frequency of the magnetic field. It is the dielectric constant in a vacuum; The calibration method uses standard samples with known dielectric constants for calibration, and establishes a quantitative relationship model between phase difference and amplitude ratio and dielectric constant, which is used to measure unknown sample matrices.

9. The method for sensing moisture in multi-depth greenhouse substrates based on dielectric spectrum according to claim 7, characterized in that, The matrix sensor adopts a three-point or ten-point calibration mechanism. At the factory, the corresponding calibration mechanism can be set via the Modbus protocol, and three-point or ten-point calibration can be selected.

Images