Self-powered multi-parameter farmland microenvironment monitoring system

Abstract

The invention discloses a self-powered multi-parameter farmland microenvironment monitoring system, and belongs to the technical field of smart agriculture and Internet of Things. The system comprises a distributed self-powered wireless sensing node array, gateway equipment and a cloud data platform. The sensing node adopts a photovoltaic and soil temperature difference synergistic hybrid energy supply module, integrates temperature, humidity, pH and other multi-parameter sensing probes, and realizes long-distance low-power-consumption transmission through an improved LoRa communication module. And the gateway is used for receiving and fusing the data and uploading the data to the cloud platform through 4G or Wi-Fi. The cloud platform can store and visualize data, spatial modeling is realized by using improved Kriging interpolation, a soil moisture content prediction model and an anomaly detection algorithm are introduced, and intelligent irrigation suggestions are generated in combination with a crop water demand model. The system has the advantages of stable energy supply, high sensing precision, strong communication anti-interference capability, intelligent data processing and the like, can operate unattended for a long time, is widely suitable for scenes of field planting, greenhouse agriculture, scientific research experiments and the like, and provides reliable support for precise and intelligent management of agricultural production.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of smart agriculture and Internet of Things technology, and particularly relates to a self-powered multi-parameter farmland micro-environment monitoring system, which is suitable for real-time data acquisition and intelligent management in the monitoring of field crops, greenhouse agriculture and ecological environment. BACKGROUND

[0002] With the development of smart agriculture, farmland micro-environment monitoring systems play an increasingly key role in fine planting management, agricultural automation and climate adaptability decision-making. Through real-time sensing and remote transmission of soil temperature and humidity, pH value and environmental temperature and other multi-parameters, the monitoring system can provide scientific data support for farmland management and promote the intelligentization and sustainability of agricultural production.

[0003] At present, common farmland monitoring systems are usually composed of wireless sensing nodes, communication networks and data processing platforms. In actual deployment and application, the following key technical challenges still exist:

[0004] 1. Limited node power supply capacity, insufficient running stability

[0005] Current large number of sensing nodes rely on disposable batteries or single photovoltaic modules for power supply. Limited by the uncertainty of the field environment, such as continuous rainy weather, more shaded areas, insufficient temperature difference and other factors, it is easy to cause power interruption or frequent disconnection. In addition, frequent battery replacement not only increases the maintenance workload, but also increases the operation and maintenance cost of the system.

[0006] 2. Sensing accuracy is easily disturbed, data reliability is not high

[0007] The soil environment of farmland has significant spatial heterogeneity and complex electrochemical properties, especially in pH detection, humidity measurement, etc. Traditional sensors often drift due to external ion concentration fluctuations, temperature changes and pollution deposition, affecting data accuracy. At the same time, most sensing nodes are single-point deployed, which is difficult to reflect the overall condition of the land.

[0008] 3. Weak anti-interference ability of communication link, weak system reliability

[0009] In the agricultural environment, traditional 2.4GHz wireless communication signals are easily affected by factors such as vegetation obstruction, soil humidity, electromagnetic interference, etc., and the transmission stability is insufficient. While some existing low-power wide-area communication technologies such as LoRa have long-distance transmission capability, the communication parameters are mostly statically configured, which is difficult to dynamically adjust according to the actual link quality, and is prone to packet loss, delay and other problems, affecting data integrity.

[0010] 4. Data processing is scattered, lacking effective fusion mechanism

[0011] The data collected by multiple nodes are heterogeneous in time, space and dimension, and it is difficult to be used for fine management and intelligent decision without effective data fusion and model support. Meanwhile, some systems lack of identification and filtering mechanism for abnormal data, which further affects the data availability.

[0012] In summary, there is an urgent need for a distributed monitoring system with high reliable energy supply capability, supporting multi-parameter high-precision sensing, with anti-interference adaptive communication mechanism, and integrating intelligent data processing capability, to realize the goal of long-term stable operation, low operation and maintenance cost and high data value of agricultural information collection and management in the field of farmland micro-environment monitoring. SUMMARY

[0013] In view of the problems of insufficient energy supply stability, low sensing precision, poor communication link reliability and limited data processing capability of existing farmland micro-environment monitoring systems, a new technical solution is proposed. The overall concept is to introduce a hybrid energy supply module into the sensing node, combined with multi-parameter high-precision sensing probe and adaptive anti-interference communication mechanism, and relying on cloud platform to realize intelligent data fusion and decision support, so as to build a low-power, long-life, high-reliable distributed farmland micro-environment monitoring system. The technical solutions of the present application will be described in detail in combination with embodiments.

[0014] In an embodiment of the present application, a self-powered multi-parameter farmland micro-environment monitoring system is provided, comprising:

[0015] Self-powered wireless sensing node array, each node comprising:

[0016] Hybrid energy supply module, composed of micro photovoltaic unit and soil temperature difference power generation unit, wherein:

[0017] The micro photovoltaic unit is a GaAs thin film battery, and the light conversion efficiency under the condition of 200 lux is not less than 28%;

[0018] The temperature difference power generation unit contains axially arranged Bi2Te3 thermoelectric arms, and the output power is not less than 3.2mW / cm 2 ;

[0019] Multi-parameter sensing module, integrating the following sensors:

[0020] MEMS temperature sensor, based on Pt1000 thin film resistor, measurement accuracy is ±0.3℃;

[0021] Interdigital capacitance humidity sensor, measurement range 0-100% RH, error is ±2%;

[0022] ISFET pH sensor, using antimony reference electrode, measurement accuracy is ±0.2pH;

[0023] The communication module adopts an improved LoRa modulation mode and supports dynamic switching of spreading factors SF7-SF12.

[0024] Further, the system further comprises a gateway device for receiving node data and uploading to a cloud platform, the gateway device comprising:

[0025] An adaptive frequency hopping controller dynamically selects a sub-channel of the 902-928 MHz frequency band according to the RSSI.

[0026] A data fusion unit for Kriging interpolation processing of spatially heterogeneous data of multiple nodes.

[0027] Preferably, the hybrid energy supply module further comprises:

[0028] A super low power consumption power management chip of model BQ25570, with a start-up voltage of 0.3V;

[0029] An energy distribution optimization algorithm, which executes the following strategies:

[0030]

[0031] Optionally, the hybrid energy supply module further comprises a radio frequency energy harvesting unit, whose working frequency band includes 902-928 MHz and 2.4 GHz, and which is automatically activated when the combined output power of photovoltaic and thermoelectric is lower than 1.5mW.

[0032] Further, the radio frequency energy harvesting unit comprises:

[0033] A dual-band rectifier antenna, a Schottky diode rectifier, a dynamic impedance matching network and an energy management module for converting radio frequency signals into direct current and supplying energy.

[0034] Preferably, the system controls the working time slot of the radio frequency rectifier through a communication scheduling mechanism to avoid conflicts with LoRa communication.

[0035] In an embodiment of the present application, the probe surface of the multi-parameter sensing module is provided with:

[0036] A hydrophobic nano-coating with a contact angle greater than 150°, the main component of which is fluorinated silicon dioxide;

[0037] A self-cleaning electrode structure that removes contaminants by periodically applying square wave pulses of -1.2V to +0.8V.

[0038] Further, the transmission protocol of the communication module comprises:

[0039] (1) Spreading factor dynamic adjustment strategy based on packet error rate (PER), the logic of which is as follows:

[0040]

[0041] (2) Time slot allocation method: dynamically allocate transmission priority based on soil moisture value of each node, nodes with moisture exceeding 80% get double transmission time slots.

[0042] In an embodiment of the present application, a farmland sensor network deployment method is provided, comprising:

[0043] Node arrangement stage:

[0044] Grid division according to soil type, clay area deployment density is 30 nodes / hectare, sandy area is 20 nodes / hectare;

[0045] Probe insertion depth is set to the main root distribution layer ±5cm, standard is 15cm;

[0046] Network self-calibration stage:

[0047] Gateway issues TDMA time slot table, time slot length is 1.3 times of data packet transmission time;

[0048] Two-point calibration is performed on the pH sensor, and pH=4.01 and pH=7.01 buffer solutions are used respectively.

[0049] In an embodiment of the present application, a LoRa signal anti-interference method is provided, comprising:

[0050] Spectrum sensing step: detect the interference intensity of 902-928MHz frequency band through fast Fourier transform (FFT);

[0051] Dynamic avoidance step: when a sub-channel RSSI<-90dBm and PER>10%, switch to a backup channel;

[0052] Transmit power optimization step: dynamically adjust the transmit power according to the following formula:

[0053] P tx = min(14dBm,-0.3×distance+20)

[0054] In an embodiment of the present application, the cloud platform performs the following tasks:

[0055] Soil moisture prediction model:

[0056] w(t+1)=0.7w(t)+0.2T(t)-0.1ETo(t)

[0057] Where w(t) is the current soil moisture, T(t) is the current air temperature, and ETo(t) is the reference crop evapotranspiration.

[0058] Outlier filtering algorithm: ±3 sigma range filtering is performed on the pH sensor data, and abnormal data is removed.

[0059] Based on the above technical scheme, the self-powered multi-parameter farmland microenvironment monitoring system of the application realizes high reliability, low energy consumption and long period distributed monitoring for multiple scenes such as field crops and greenhouse agriculture through integration of hybrid energy supply mechanism, multi-parameter high-precision sensing technology, adaptive LoRa communication protocol and intelligent data processing platform, effectively solving the core problems of existing farmland environment monitoring system in energy supply, measurement accuracy and communication reliability.

[0060] Compared with the prior art, the application has the following technical effects:

[0061] 1. The application adopts a hybrid energy supply design of photovoltaic and soil temperature difference cooperation, wherein the GaAs thin film battery has weak light high-efficiency conversion capability, and the thermoelectric unit continuously outputs electric energy by utilizing the soil-air temperature difference, and is dynamically scheduled through the BQ25570 ultra-low power supply management chip and energy distribution algorithm. Compared with the traditional node design relying on dry batteries or single photovoltaic source, the system can realize self-sustained operation under adverse conditions such as rain and shade, and the energy supply stability reaches 100%, which significantly reduces the operation and maintenance frequency and cost.

[0062] 2. The sensing node of the application integrates three types of high-precision sensors of temperature, humidity and pH based on MEMS technology, and through the innovative "hydrophobic nano coating + self-cleaning electrode" structure, effectively prolongs the working life of the sensor in the contaminated medium such as soil, and improves the data consistency. Experimental results show that the pH measurement standard deviation of the system can be controlled within ±0.18, which is more than 50% higher than the commercial scheme.

[0063] 3. The LoRa communication module of the application supports adaptive adjustment of the spreading factor from SF7 to SF12, dynamically configures the communication parameters through RSSI sensing and PER feedback, and combines the time slot priority mechanism to ensure the priority upload of the key node in high humidity. The anti-interference algorithm further reduces the packet loss rate from 22.7% to 8.3% under long-distance (>500m) transmission conditions through FFT detection and transmission power optimization strategy, which greatly improves the data integrity.

[0064] 4. The application proposes a node layout strategy and self-calibration process based on soil type, which supports differential deployment in different crops and land types. The standardized calibration process can significantly improve the alignment of pH and humidity data in the spatial dimension, providing a basis for realizing spatial accurate restoration of farmland microenvironment.

[0065] 5. The cloud platform of this invention introduces Kriging interpolation enhanced with crop growth models to generate spatial distribution maps. It combines this with LSTM anomaly detection algorithms, soil moisture prediction models, and irrigation decision logic to achieve a closed-loop chain from data collection to intelligent decision-making. The platform supports integration with third-party agricultural information systems, providing agricultural managers with visualized and intelligent decision support.

[0066] In summary, this invention, by systematically integrating low-power sensor technology, hybrid energy acquisition mechanism, anti-interference remote communication strategy, and cloud-based intelligent algorithm, achieves multi-dimensional and high spatiotemporal resolution data acquisition of farmland microenvironment while ensuring long-term stable operation. This improves the precision and intelligence of agricultural production and has broad prospects for promotion and industrial application value. Attached Figure Description

[0067] Figure 1 This is a schematic diagram of the network topology of the self-powered multi-parameter farmland microenvironment monitoring system of the present invention, used to illustrate the communication relationship between the sensor nodes, gateway devices and cloud platform in the system.

[0068] Figure 2 This is a schematic diagram of the structure of the hybrid power supply module in the self-powered wireless sensing node, showing the connection relationship between the micro photovoltaic unit, the soil temperature difference power generation unit, the radio frequency energy harvesting module, and the power management chip.

[0069] Figure 3 This is a schematic cross-sectional view of the structure of a multi-parameter sensing probe, showing the layered distribution of the MEMS temperature sensor, interdigitated capacitive humidity sensor, ISFET pH sensor, and their surface hydrophobic coating and self-cleaning electrode structure.

[0070] Figure 4 This is a comparison chart of the packet loss rate of the communication scheme of the present invention and the traditional communication scheme at different communication distances, used to illustrate the superiority of the adaptive LoRa communication mechanism in terms of anti-interference performance.

[0071] Figure 5 It is a block diagram of the radio frequency energy harvesting module and its energy coordination logic diagram with the hybrid power supply module, showing the collaborative working relationship between the dual-frequency rectifier antenna, rectifier circuit, dynamic impedance matching network and energy management module. Detailed Implementation

[0072] To make the objectives, technical solutions, and beneficial effects of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the described embodiments are only a part of the implementation methods of this invention, and not all of them. Other implementation methods obtained by those skilled in the art based on the embodiments of this invention without creative effort should all fall within the protection scope of this invention.

[0073] In the following specific implementation, it will be combined with Figures 1 to 4 This paper elaborates on the system structure, power supply module, sensing probe, communication protocol, deployment and calibration process, and cloud platform application of the present invention, and provides experimental verification results to facilitate those skilled in the art to understand and implement the present invention.

[0074] It should be understood that the following embodiments are for illustrative purposes only and do not limit the scope of protection of the present invention.

[0075] I. System Overall Structure

[0076] In one embodiment of the present invention, a self-powered multi-parameter farmland microenvironment monitoring system is provided, the system structure of which is as follows: Figure 1 As shown. The system adopts a distributed architecture and mainly consists of multiple self-powered wireless sensor nodes deployed in farmland, gateway devices set at the edge, and a remote cloud data platform. The three work together through a wireless communication network to form a complete agricultural information monitoring closed loop from sensing, transmission to decision control.

[0077] During system operation, sensor nodes are distributed in a grid pattern within the farmland area to collect various key agricultural environmental parameters in real time, including soil temperature, humidity, and pH value. Each sensor node integrates a hybrid power supply module, a multi-parameter sensing module, and a communication control module, possessing independent operation and self-scheduling capabilities. The hybrid power supply module utilizes a dual-energy path based on GaAs thin-film photovoltaic cells and Bi2Te3 thermoelectric power generation components, supplemented by a radio frequency energy harvesting module as an emergency power supply. This ensures stable node operation under various weather conditions, eliminating the need for external power sources or frequent battery replacements, significantly improving the system's deployment flexibility and operational efficiency.

[0078] The multi-parameter sensing module adopts an integrated probe structure, embedding a MEMS temperature sensor, an interdigital capacitive humidity sensor, and an ISFET pH sensor. It can stably monitor soil environmental indicators for a long time, and enhances the reliability and measurement consistency of the probe in high humidity and high pollution environments through hydrophobic nano-coating and self-cleaning electrode structure.

[0079] Each sensor node establishes a wireless connection with the gateway device through an improved LoRa communication module. The LoRa module supports anti-interference mechanisms such as adaptive spreading factor (SF), RSSI-sensing frequency hopping, and time slot scheduling, achieving low-power, long-distance, and highly stable communication performance in the complex electromagnetic environment of farmland. Data collected by the nodes is transmitted to the gateway device via a wireless link within a set period, completing the first stage of data transmission.

[0080] The gateway device is located at the edge of the field or in the farm management center. Its main function is to establish a data bridge between the sensor network and the cloud platform, and to perform functions such as data aggregation, preliminary fusion, and remote forwarding. In this embodiment, the gateway has a built-in frequency hopping controller and a data fusion unit. It can dynamically select the optimal communication channel based on the strength of the signals uploaded by multiple nodes, and perform kriging interpolation processing on spatially heterogeneous data to output a high-resolution environmental information map with spatial continuity. The processed data is uploaded to the cloud via a 4G module or Wi-Fi network, supporting interruption retransmission and network anomaly buffering to ensure the integrity and continuity of data transmission.

[0081] The cloud platform serves as the system's data processing and intelligent decision-making center, providing a range of services including data reception, storage, analysis, modeling, and visualization. The platform can perform structured archiving and parameter extraction of uploaded data, supporting soil moisture prediction based on time-series models, anomaly identification based on statistical rules, and intelligent irrigation recommendations based on agricultural knowledge bases. The platform also supports graphical displays and user-interactive operations. Managers can view real-time environmental changes in various areas of the field via a web interface or mobile application, receive system-pushed warnings and irrigation recommendations, and remotely configure sensor node sampling cycles, transmission parameters, and calibration commands, enabling intelligent farmland environmental management under unattended conditions.

[0082] Through the above-described structure and process design, the system of this invention can achieve a closed-loop operation of the entire process, from multi-parameter soil sensing, wireless data acquisition, edge processing to cloud fusion and intelligent feedback. Its architecture is highly modular, scalable, and adaptable, making it suitable not only for open-field planting but also for greenhouse agriculture, high-standard farmland, and scientific research sites. It is particularly well-suited for agricultural IoT applications with stringent energy supply requirements, complex communication environments, and strong data analysis needs, demonstrating excellent practicality and promotional value.

[0083] II. Hybrid Energy Supply Module Structure

[0084] In one embodiment of the present invention, to address the problem of unstable energy supply faced by sensor nodes operating long-term in farmland environments, the self-powered wireless sensor node is equipped with a compact, energy-efficient, and highly adaptable hybrid power supply module, the structure of which is as follows: Figure 2As shown, this module integrates a multi-path micro-energy harvesting unit, a power management unit, and an energy dispatch control unit, constructing an autonomous power supply system with triple energy input capabilities. It can achieve dynamic energy acquisition and efficient conversion under different natural conditions, thereby ensuring the continuous and stable operation of the sensing node in various environments.

[0085] The hybrid power supply module includes three independent but synergistic energy input channels: a photovoltaic energy harvesting path, a thermoelectric energy harvesting path, and a radio frequency energy harvesting path. The photovoltaic power supply path utilizes GaAs (gallium arsenide) thin-film solar cells, which possess high photoelectric conversion efficiency and excellent low-light response capabilities. Even in low-illuminance environments of 200 lux, they can maintain a conversion efficiency of over 28%, making them particularly suitable for solar energy utilization inside greenhouses or during cloudy / rainy weather. This photovoltaic module enhances incident light capture capability through a microlens array and anti-reflective film structure, making it suitable for all-weather deployment.

[0086] The thermoelectric energy harvesting path, connected in parallel with the photovoltaic path, is based on a Bi2Te3 (bismuth, antimony, tellurium) thermoelectric material to construct an axially arranged thermoelectric arm structure. The hot end faces into the soil, and the cold end faces the air. Driven by the diurnal temperature difference, the electron flow is utilized. When the temperature difference ΔT ≥ 5K, its output power per unit area can reach 3.2 mW / cm². 2 A thermally insulating filler and a heat dissipation channel are installed between the hot and cold ends to maintain heat flow stability and improve energy conversion efficiency. The thermoelectric path is particularly suitable for energy compensation at night, on rainy days, or during periods of insufficient sunlight in winter.

[0087] To address extreme weather conditions or challenging environmental conditions such as continuous low light levels and extreme temperature differences, this invention further introduces a radio frequency (RF) energy harvesting path as an emergency power supply method. This RF path includes a dual-band rectifier antenna, a Schottky diode rectifier, a dynamic impedance matching network, and an energy management module. It can simultaneously receive RF signals in the 902–928MHz and 2.4GHz frequency bands, adapting to current mainstream LoRa and Wi-Fi signal sources. When the combined output power of the photovoltaic and thermoelectric paths falls below 1.5mW, the system automatically activates the RF energy harvesting module to harvest low-power RF energy from the environment non-contactly and convert it into DC power to maintain basic node operation.

[0088] The input terminals of all three energy harvesting paths are connected to a power management unit, whose core is the BQ25570 ultra-low-power power management chip. This chip features an ultra-low cold-start capability of 0.3V, making it suitable for micro-power energy environments. It also integrates a maximum power point tracking (MPPT) algorithm, which can adjust the operating point in real time to improve conversion efficiency when the photovoltaic path is active. Furthermore, the chip supports dual-input channel management, energy path priority configuration, energy storage capacitor scheduling, and undervoltage protection, enabling centralized management and intelligent scheduling of various input energy sources.

[0089] In a preferred embodiment of the present invention, the system dynamically determines and switches between photovoltaic and thermoelectric paths through an energy allocation optimization algorithm embedded in the firmware. The scheduling logic is as follows:

[0090]

[0091]

[0092] Here, P_solar and P_therm represent the currently collected photovoltaic power and thermoelectric power, respectively. When the output of the thermoelectric path is higher than that of the photovoltaic, the system will actively disconnect the photovoltaic DC-DC boost path to avoid energy reinjection and energy consumption through non-optimal paths, ensuring that the currently available energy enters the energy storage unit or direct power supply module through the optimal path.

[0093] Furthermore, to enhance the system's stability during long-term operation, the power management module incorporates an energy gating mechanism and state-maintaining logic. The system periodically assesses the energy storage unit's state of charge, voltage change rate, and external environmental trends, implementing hysteresis control on the energy path to prevent frequent switching that could cause system oscillations, thereby improving the stability and lifespan of the power supply process.

[0094] Through the aforementioned structural design and strategic control, the hybrid power supply module of this invention can autonomously adapt and dynamically switch under different light intensities, temperature differences, and radio environments, significantly enhancing the system's energy acquisition capability and long-term stable operation capability in complex farmland environments. Compared to traditional single-path power supply sensor node structures, this invention significantly reduces the probability of node disconnection due to power supply interruptions, while avoiding the labor costs and environmental burden caused by frequent battery replacements, demonstrating high engineering practicality and deployment feasibility.

[0095] III. Sensor Probe Structure

[0096] In one embodiment of the present invention, to achieve long-term stable monitoring of multi-dimensional parameters of farmland soil, the sensing node is equipped with a multi-parameter sensing probe with high structural integration, high measurement accuracy, and strong environmental adaptability, the structure of which is as follows: Figure 3 As shown, this probe adopts a vertically integrated packaging method, internally integrating three types of sensing units: temperature, humidity, and pH. Its external structure incorporates a hydrophobic coating and a self-cleaning electrode design, enabling it to adapt to the complex conditions of high humidity, high salinity, and high corrosiveness in farmland environments, achieving data acquisition in long-term unattended operation.

[0097] The sensing probe is mounted on a cylindrical SiO2 ceramic substrate, which possesses excellent electrical insulation, mechanical strength, and thermal stability, making it an ideal platform for constructing multiple sensing functional layers. Multiple sensing layers are sequentially deposited and patterned on this substrate using MEMS micro / nano fabrication processes, specifically including the following components:

[0098] First, at the bottom of the probe is a Pt1000 thin-film temperature sensor. This sensor uses magnetron sputtering technology to deposit platinum metal onto an insulating layer to form a thin-film resistor, exhibiting excellent linear response characteristics and temperature stability, with a measurement accuracy of ±0.3℃. Its short response time and low signal drift make it suitable for high-frequency sampling tasks, accurately reflecting soil temperature change trends.

[0099] An interdigitated capacitive humidity sensing structure is placed above the temperature sensing layer. This structure forms a capacitor array sensitive to changes in air / soil moisture by arranging metal finger electrodes in an alternating pattern on a plane. When soil moisture increases, causing a change in the dielectric constant between the electrodes, the sensor's output capacitance value also changes. The system can calculate the relative humidity by measuring this value. Compared with traditional conductivity-based humidity sensors, the interdigitated capacitive structure exhibits stronger anti-interference capabilities and lower sensing drift characteristics in strongly electrolyte environments such as high-salt and acidic soils.

[0100] Following this is the ISFET (Ion-Sensitive Field Effect Transistor) pH sensor. This sensor employs a structure combining an antimony-based reference electrode with an Al2O3 gate dielectric layer, enabling highly sensitive responses to changes in hydrogen ion concentration. To improve measurement consistency under field conditions, the system also incorporates a temperature compensation module, introducing the following correction model for data fitting:

[0101] pH corrected =pH raw +β(T-25)

[0102] Where T is the soil temperature value collected by the probe, and β is the temperature drift coefficient (empirical value is about 0.02). This compensation method can control the standard deviation of pH measurement under different seasons and day and night conditions within ±0.18.

[0103] To address the issue of probe performance degradation due to contamination, scaling, and microbial adhesion during long-term burial, this invention introduces a dual-layer protection design on the probe surface: Firstly, a nano-hydrophobic coating composed of fluorinated silica is uniformly coated on the outer shell surface of the sensing probe. This coating is formed using plasma vapor deposition (PECVD) and has a surface contact angle greater than 150°, exhibiting excellent resistance to mud and water adhesion and waterproofing performance. Secondly, a self-cleaning electrode structure is set in the sensing electrode area. The node master control chip periodically controls the application of a square wave pulse signal from -1.2V to +0.8V to the electrode, thereby disrupting the binding of deposited ions and contaminant particles through electrochemical stripping, achieving automatic decontamination without the need for manual maintenance.

[0104] In terms of mechanical packaging, the probe adopts an integral injection-molded structure, and the shell material is made of high-strength polycarbonate (PC) engineering plastic, which has good corrosion resistance and waterproof sealing performance, and the protection level can reach IP68, meeting the requirements for year-round buried operation. The diameter of the entire probe is less than 60mm, and the length is controlled within 120mm, which can be easily inserted into the cultivated layer or the crop root distribution area. The standard installation depth is 15cm±5cm.

[0105] Each sensing unit inside the probe is connected to the node's main control module via flexible wiring. All sensing signals are conditioned and acquired using a unified ADC channel. The system supports time-sharing power-on, low-power standby, and fast wake-up mechanisms to reduce energy consumption and extend system lifespan. According to test results, under standard installation conditions, the annual maintenance frequency of the sensing probe is less than once, and the stable operation cycle can exceed 18 months, making it suitable for large-scale deployment and unattended scenarios.

[0106] Through the above structural design and material optimization, the multi-parameter sensing probe provided by this invention not only meets the high standards of agriculture in terms of measurement accuracy, but also has significant advantages in terms of structural durability, environmental adaptability and operation and maintenance costs. It is an important supporting component for realizing high-density and high-consistency agricultural environmental monitoring.

[0107] IV. Communication Protocols and Anti-interference Mechanisms

[0108] In one embodiment of the present invention, in order to ensure stable data upload by the sensing node in open, complex, and easily attenuated environments such as farmland, the system adopts a low-power wide-area communication protocol based on LoRa (Long Range) modulation between the node and the gateway, and further designs several adaptive enhancement mechanisms on this basis to effectively improve the reliability, anti-interference ability and power consumption control level of the system communication.

[0109] like Figure 4As shown, under different transmission distance conditions, the communication mechanism adopted in this invention is significantly better than the traditional LoRa static configuration scheme in terms of packet loss rate control, especially at communication distances of more than 600 meters, where its data integrity is significantly improved.

[0110] LoRa, as a spread spectrum modulation method, inherently possesses the characteristics of low power consumption, strong penetration, and high sensitivity, making it suitable for communication needs in agricultural scenarios where sensor node power consumption is limited, distances are long, and environmental obstructions are severe. However, traditional LoRa communication protocols mostly employ static spreading factor configuration, which cannot be adjusted in real time according to the actual quality of the communication link. This leads to problems such as data packet loss, communication delays, or node disconnections under conditions such as long distances, strong interference, or drastic humidity changes.

[0111] To overcome the aforementioned shortcomings, this invention introduces an adaptive spreading factor adjustment mechanism based on the LoRa physical layer protocol. Specifically, each node monitors its Packet Error Rate (PER) with the gateway in real time during data upload and dynamically adjusts its spreading factor (SF) based on the PER value. When the PER value remains above 15%, the node automatically increases its SF level (e.g., from SF7 to SF9) to enhance signal processing redundancy and anti-interference capabilities; conversely, when the PER value stabilizes below 5%, the system gradually reduces the SF to improve transmission rate and spectrum utilization efficiency. This adjustment strategy achieves an adaptive balance between communication quality and power consumption, adapting to various influencing factors such as sunlight, humidity, crop shading, and changes in the electromagnetic environment.

[0112] Furthermore, to further reduce interference sensitivity within the communication frequency band, this invention deploys an RSSI-aware frequency hopping controller at the gateway. This module periodically scans multiple sub-channels within the 902–928MHz frequency band, measuring their Received Signal Strength (RSSI) and background noise levels. When the current communication channel's RSSI is found to be below -90dBm or the corresponding PER exceeds 10%, the system actively switches to a backup channel with lower interference. This mechanism effectively avoids communication instability issues caused by channel congestion, changes in interference sources, and other factors in actual deployments with fixed frequency configurations.

[0113] Regarding the MAC layer protocol, this invention also introduces a time-slot priority scheduling strategy based on farmland moisture thresholds. During the initialization phase, the gateway issues a time-slot scheduling table based on TDMA (Time Division Multiple Access), and each sensor node completes data upload within a specified time window. To ensure priority transmission of data in high-humidity areas or areas with abnormal soil moisture, the system dynamically adjusts the communication priority of each node based on the soil moisture values ​​collected. When a node detects a humidity value exceeding 80%, the system allocates double the upload time slots to it, thereby ensuring the timeliness and completeness of critical agricultural data and improving the real-time performance of intelligent irrigation decisions.

[0114] To further control power consumption and reduce energy waste caused by simultaneous communication of multiple nodes, this invention also proposes a dynamic transmission power adjustment method based on distance estimation. This method controls the transmission power of the LoRa module according to the following linear model, based on the relative distance from the node to the gateway:

[0115] P tx =min(14dBm,-0.3×distance+20)

[0116] Where distance represents the estimated distance (in meters) between the node and the gateway, P tx The calculated transmit power value (in dBm) is used to limit the maximum output power to no more than 14 dBm to meet radio transmit power regulations. This strategy can effectively reduce node power consumption and extend the overall system uptime while ensuring reliable communication.

[0117] Experimental data shows that in a typical field application scenario with a transmission distance of 600 meters, the average packet loss rate of nodes using the communication mechanism of this invention is controlled within 10%, while the packet loss rate of traditional LoRa fixed configuration schemes is generally higher than 25%. At the same time, the average transmit power required by the node for the communication strategy of this invention is about 72% of that of traditional schemes, which significantly extends the system life without sacrificing communication quality.

[0118] In summary, the communication protocol and its series of adaptive enhancement mechanisms adopted in this invention can dynamically adjust key communication parameters according to the actual changes in the farmland communication environment, effectively improving data integrity, communication robustness and system energy efficiency. It is especially suitable for IoT application scenarios in agricultural environments with dense node deployment, complex interference and difficult operation and maintenance.

[0119] V. Coordinated Mechanism of Radio Frequency Energy Harvesting and Power Supply

[0120] In one embodiment of the present invention, to further enhance the system's energy security under extreme weather conditions and improve the long-term operational stability of the sensing nodes, a radio frequency energy harvesting module is integrated into the hybrid power supply module, the structure and operation of which are as follows:Figure 5 As shown, this module serves as an auxiliary compensation mechanism for the photovoltaic and thermoelectric energy harvesting paths. It is mainly used to provide low-power energy input in extreme environments such as continuous rain, severe shading, or insufficient diurnal temperature range, ensuring uninterrupted supply of the minimum power consumption required for node operation.

[0121] like Figure 5 As shown, the radio frequency energy harvesting module includes the following components:

[0122] RF rectifier antenna unit;

[0123] High-frequency rectifier circuit (Schottky diode array);

[0124] Dynamic impedance matching network;

[0125] Energy management and boost module;

[0126] System-level scheduling and control logic interface.

[0127] The rectifier antenna unit adopts a dual-band design, resonating in the 902–928MHz and 2.4GHz bands respectively, to adapt to the environmental radio frequency energy generated by mainstream LoRa base stations and Wi-Fi routers. The antenna structure employs a multi-layer PCB resonant ring design, combined with dielectric loading technology to enhance miniaturization and integration capabilities. In actual deployment environments, this antenna can stably receive signals with an effective radiated power in the air of no less than -10dBm, and possesses good directivity and gain characteristics.

[0128] After the radio frequency (RF) signal is received by the antenna, out-of-band noise is filtered out by a front-end bandpass filter before being sent to a high-frequency rectifier circuit for rectification and conversion. The core of this rectifier module is a Schottky diode array, which, due to its extremely low forward voltage drop and high-speed response characteristics, is suitable for DC-DC conversion of low-power, high-frequency RF signals. To improve rectification efficiency, the rectifier employs a multi-stage voltage multiplier topology, enabling it to output DC voltages above 0.3V under extremely low signal input power conditions, providing a voltage foundation for subsequent energy processing.

[0129] The rectified voltage is then input to a dynamic impedance matching network. Because the strength and frequency characteristics of RF signal sources fluctuate under different environments, impedance mismatch between the antenna and the rectifier circuit will severely reduce energy transfer efficiency. This embodiment employs a tunable network composed of a variable inductor and a varactor diode. By detecting the input power and output voltage, the matching parameters are adaptively adjusted in real time to achieve optimal impedance matching between the antenna and the rectifier. This mechanism can effectively improve rectification efficiency by more than 15%, and is particularly suitable for open environments where the signal source changes dynamically.

[0130] The electrical energy output from the impedance network enters the energy management and boost module, which internally includes a boost DC-DC converter circuit, a charge pump, an energy buffer capacitor, and an energy state detection circuit. The system can dynamically determine whether this energy enters the energy storage unit or directly enters the main power supply circuit based on the current node operating status. When the main power supply path is insufficient due to low light or temperature differences, the RF path will provide stable low-power (0.1–0.8mW) continuous replenishment capability, effectively supporting the node's minimum operating power consumption and data reporting function.

[0131] To achieve intelligent scheduling and energy coordination between the RF module and the two main energy paths of photovoltaic and thermoelectric power, this invention designs a unified energy control strategy. The system periodically monitors the current output power of each power supply path through the main control MCU and determines the energy switching behavior according to the following three-level priority strategy:

[0132] 1. First-level power supply: Under good solar irradiance (P_solar≥1.5mW), the photovoltaic path is activated first, and the MPPT algorithm is started for maximum power point tracking;

[0133] 2. Secondary power supply: When sunlight is insufficient but the diurnal temperature range is significant (ΔT≥5K), switch to the thermoelectric path as the main power source and shut down the photovoltaic path DC-DC converter to avoid energy backflow;

[0134] 3. Third-level power supply: When the combined output power of the photovoltaic and thermoelectric paths is less than 1.5mW and the node power is lower than the maintenance threshold (e.g., V_batt<2.2V), the radio frequency energy harvesting path is activated to provide compensation power.

[0135] To prevent the high-frequency rectification process on the spectrum from interfering with LoRa communication during RF path activation, this invention introduces an energy-coordinated control mechanism based on communication scheduling. Specifically, the system allows the RF rectification module to operate only during non-communication window periods, according to the TDMA communication time slot table issued by the gateway; once the LoRa communication window is entered, the excitation path of the RF rectifier is immediately shut down to prevent high-frequency rectification noise from interfering with signal modulation and identification or causing communication failure.

[0136] Field test results show that in typical southern rice-growing areas, under seven consecutive days of cloudy and rainy weather, the online rate of traditional photovoltaic-thermal power dual-mode nodes is approximately 68%, while the online rate of nodes with integrated radio frequency energy harvesting modules increases to 92%, and the data upload success rate improves by 24%. This module is particularly suitable for deployment in areas with significant shading (such as greenhouse steel structure areas), continuous low light (such as cloudy seasons), or areas with high-density vegetation shading, effectively filling power supply blind spots in the main energy path and improving system robustness and unattended operation capabilities.

[0137] In summary, the radio frequency energy harvesting and power supply coordination mechanism described in this invention significantly improves the energy security level of nodes in extreme environments without increasing system power consumption by introducing a controllable and efficient radio frequency energy capture structure and dynamic scheduling algorithm. It forms a complementary and redundant triple power supply system with photovoltaic and thermoelectric paths, which is a key supporting technology for the long-term stable deployment of agricultural monitoring systems.

[0138] VI. Sensor Network Deployment Methods and Calibration Procedures

[0139] In one embodiment of the present invention, to achieve high-density, spatially continuous sensing of multi-parameter environmental information in farmland and ensure the deployment effect and measurement accuracy of the system in actual agricultural production environments, the system further provides a sensor network deployment method and node calibration process. This method optimizes deployment by region for different soil types, crop distributions, and terrain conditions. Combining the structural characteristics of the sensors themselves and data consistency requirements, it establishes standardized deployment depth, density, and calibration parameter initialization procedures, thereby establishing a structurally sound and synchronously accurate wireless monitoring network.

[0140] Before system installation, the monitoring area is first divided into grids based on soil texture, topographic elevation, and crop type. Preferably, the target monitoring area is divided into 50m x 50m cells. In areas with loam or sandy soil, the sensor node deployment density can be set to 20 nodes per hectare; while in areas with high clay content, uneven moisture distribution, or mixed crop varieties, the deployment density is increased to 30 nodes per hectare. During node deployment, areas prone to environmental interference, such as agricultural machinery paths, drainage ditches, and building shadows, should be avoided as much as possible to ensure the representativeness and validity of the collected data.

[0141] After the nodes are deployed, the probes are vertically inserted into the soil according to their length and structure, and the location of the sensing layer. The insertion depth is controlled at 15cm ± 5cm to match the distribution layer of the main root system of crops. To ensure sufficient contact between the probe and the soil, the surrounding soil should be lightly compacted after insertion to avoid air gaps that could cause measurement errors. For areas with high pH monitoring requirements, it is recommended to set up at least one high-precision pH node in each grid cell to facilitate subsequent soil improvement and nutrient management support.

[0142] After all nodes have completed their physical installation, the system initialization phase begins. This phase mainly includes two parts: the sensor calibration process and the network communication synchronization process.

[0143] For sensor calibration, preferably, the system employs a two-point solution calibration method for the pH sensor. Specifically, the probe is first cleaned and then sequentially inserted into standard buffer solutions at pH = 4.01 and pH = 7.01, respectively. Stable output voltage values ​​are collected, and a calibration parameter table is automatically generated based on the response curve. The system writes this parameter table into the node's internal Flash memory for subsequent correction processing of actual soil pH data. For higher accuracy, it can be extended to three-point calibration (pH = 4.01, 6.86, 9.18).

[0144] For temperature and humidity sensors, since they have already undergone factory-level calibration, on-site calibration is generally not required unless there are special environmental requirements. However, after installation, the system will perform a rapid environmental balancing sampling procedure, continuously measuring background temperature and humidity data for 30 minutes to establish a local initial offset reference value for the sensor, which is used to detect individual inaccuracies or installation deviations.

[0145] After completing the calibration process, the system enters network-level initialization. The gateway device, acting as the master control node, broadcasts initialization beacon information to each sensor node, including: network ID, working channel number, TDMA communication time slot configuration table, node number allocation strategy, and heartbeat cycle parameters. Upon receiving the beacon information, each node selects its reporting frequency based on its sampling period and power consumption strategy, completes communication pairing and time synchronization, and officially enters the operational state.

[0146] To improve deployment efficiency, the system also provides a mobile deployment assistant app that can communicate temporarily with nodes via Bluetooth to enable rapid functional testing, signal strength assessment, parameter download, and tag location during deployment, assisting operators in completing batch deployment tasks in the field.

[0147] According to statistics from actual operation, after adopting the above deployment methods and calibration procedures, more than 90% of the system nodes can enter a stable operating state after the first installation, the data error is controlled within ±5%, and the average communication establishment time does not exceed 120 seconds, which significantly improves the efficiency of large-scale deployment and the level of data consistency of the agricultural Internet of Things system.

[0148] In summary, the sensor network deployment method and calibration process provided by this invention establish a deployment mechanism suitable for high-density node systems in farmland environments, based on soil adaptability, rational layout, parameter consistency, and network synchronization. It has clear engineering operability and mass production promotion value, and is a key foundational link for the system of this invention to achieve accurate monitoring and intelligent response.

[0149] VII. Data Processing and Intelligent Decision Support Module of Cloud Platform

[0150] In one embodiment of the present invention, the self-powered multi-parameter farmland microenvironment monitoring system collects environmental parameters through sensor nodes deployed in the farmland and uploads the data to a gateway device via a LoRa communication module. The data is ultimately aggregated to a cloud platform for unified processing and analysis, thereby enabling functions such as environmental perception, trend prediction, anomaly warning, and irrigation decision support in agricultural production management. This cloud platform, serving as the system's remote intelligent control center, has a data processing flow comprising multiple functional modules, including data reception, parsing and storage, quality control, spatial modeling, intelligent decision reasoning, and visualization output, forming a complete "perception-analysis-decision-feedback" closed loop.

[0151] During the data reception phase, the cloud platform continuously receives monitoring data packets uploaded from the edge gateway via 4G or Wi-Fi interfaces. The system adopts a unified data packet structure definition, including multiple fields such as node number, sampling timestamp, GPS location information, temperature, humidity, pH value, battery voltage, and communication quality parameters (such as RSSI and SNR). After the data is transmitted to the platform, it first undergoes structured parsing and caching processing before being efficiently stored in a time-series database.

[0152] To ensure the accuracy of data analysis, the platform is equipped with an automated data quality control module, employing a multi-layered verification mechanism based on sliding window mean detection, Z-score outlier identification, and logical rule filtering. For data points that are duplicated, missing, exhibit abnormal fluctuations, or have disordered timestamps, the system can automatically mark them as pending verification and determine whether to correct or remove them based on the preceding and following data trends, minimizing data contamination caused by environmental interference or sensor errors.

[0153] After data quality control, the system enters the spatial modeling phase. Based on the spatial coordinate information of each node, and combined with the Kriging interpolation algorithm, the platform performs spatial continuous modeling of environmental parameters such as temperature, humidity, and pH. This modeling process can not only generate soil moisture heat maps and pH distribution maps at any given time, but also serve as the basis for subsequent variable prediction and control strategy generation. The specific modeling functions are as follows:

[0154]

[0155] Where Z(x0) is the parameter value of the location to be predicted, Z(x i ) represents known node data, λ i The weighting coefficients are determined by the covariance function between nodes, and n is the number of neighboring nodes involved in the calculation. This method enables high-precision data field recovery in sparsely covered areas.

[0156] After the model was built, the platform further provides an intelligent decision support module. This module addresses the most critical issue in agricultural irrigation management—water regulation—by designing an intelligent irrigation suggestion generation algorithm based on a soil water holding capacity model and crop water demand curves. The system dynamically calculates whether there is a risk of water shortage in the target area based on historical humidity trends, current soil moisture levels, and crop type, and proposes irrigation suggestions accordingly. Preferably, it can be corrected by incorporating weather forecast information to avoid erroneous activation of irrigation devices before rainfall, thus conserving water resources.

[0157] In terms of pH management, the platform has established a database of suitable pH models for crops and, combined with long-term monitoring data, identifies soil acidification risk trends or local high-alkalinity anomalies. It then provides users with suggestions for adjusting acidity (e.g., by applying lime) or alkalinity (e.g., by applying ammonium sulfate), enabling early intervention in soil acid-base imbalance.

[0158] All analysis results are ultimately displayed graphically on the platform's front end. Users can view multi-parameter variation curves, heat map distribution, anomaly alerts, and system operation status for each plot via a web page or mobile app. The platform also supports customized report export, historical data comparison, irrigation plan settings, and remote parameter distribution, facilitating data-driven precision field management for agricultural managers.

[0159] Furthermore, the platform is scalable and has the ability to connect with other systems, supporting data interconnection and command linkage with third-party agricultural systems (such as agricultural big data platforms, integrated water and fertilizer control systems, and agricultural machinery collaborative platforms) through API interfaces, thereby improving the level of intelligent agricultural management and control.

[0160] In summary, the cloud platform described in this invention not only possesses the ability to structure and process multi-parameter environmental data, but also integrates multiple functional modules such as spatial modeling, intelligent decision-making, and visual management. It provides agricultural production with complete support from data collection to scientific decision-making, greatly improving the efficiency, accuracy, and automation level of farmland management, and is an important component in building smart agricultural infrastructure.

[0161] VIII. Performance Verification and Typical Application Examples

[0162] In one embodiment of the present invention, in order to verify the long-term stability, sensing accuracy and energy-saving communication capabilities of the self-powered multi-parameter farmland microenvironment monitoring system under complex agricultural environments, the applicant conducted field deployment tests in a typical rice-growing area in the south and a greenhouse vegetable base in the north, and systematically compared and verified the key performance indicators.

[0163] In a southern rice paddy experimental area, this system selected a 12-hectare farmland area and deployed 240 self-powered sensor nodes at a density of 20 nodes per hectare. Each node integrates probes for three parameters: temperature, humidity, and pH. Data acquisition and transmission are performed using the hybrid power supply module described in this paper and the LoRa communication mechanism. All nodes are connected to the cloud platform through a central gateway. The system's operation cycle covers key agronomic stages such as spring plowing, transplanting, greening, heading, and grain filling, and it has been running continuously for more than 7 months without power supply replacement or manual maintenance.

[0164] During the test period, the node online rate remained above 95%, with an average of 22 valid data packets reported per day and a data collection interval of 1 hour. To verify the performance of the power supply module, the system maintained basic communication functions even under continuous rainy weather conditions (cumulative 5 days of sunshine less than 100 lux), and the minimum voltage did not fall below the node operating threshold of 2.1V, verifying the energy resilience of the three-source power supply system.

[0165] Regarding communication performance, the LoRa adaptive mechanism described in this invention was compared with the traditional fixed spreading factor configuration. At a node distance of 600 meters, the traditional solution had a packet loss rate of 27.4%, while the average packet loss rate of this invention was controlled below 8.1%, with no communication interruptions or dropped connections. System energy consumption test results show that the average daily power consumption per node in this invention is approximately 16.5 mWh, which is about 32% lower than existing systems relying on photovoltaic paths.

[0166] In deployment tests at a northern greenhouse vegetable base, to verify the system's effectiveness in structurally obstructed environments, 40 sensor nodes were deployed within a 1200-square-meter glass greenhouse. The testing focused on the radio frequency (RF) power supply path and communication anti-interference capabilities. Under typical low-light conditions in winter, light intensity was often below 150 lux, the thermoelectric temperature difference was less than 2K, and the main energy path output power was consistently below 1.0mW. The RF energy harvesting module integrated in this invention maintained basic node operation in this environment. Combined with an intelligent scheduling mechanism, it ensured that all nodes completed an average of 12 reporting tasks per day, with data integrity exceeding 97%. After long-term continuous testing, the system operated stably for over 200 days in greenhouse obstruction, high humidity, and high salt spray environments without sensor failure or manual maintenance.

[0167] In a typical application, the system of this invention has been deployed on a large scale in a smart farm in Suzhou, Jiangsu Province, for environmental monitoring and intelligent irrigation control in the facility vegetable growing area. After deployment, by linking with soil water potential sensors and irrigation control valves, automatic drip irrigation on / off control was achieved, increasing irrigation efficiency by 26% and reducing unit water consumption by 18%. Simultaneously, historical pH monitoring data assisted the farm in completing three local lime application improvement operations within a year, alleviating the problem of yellowing cucumber leaves caused by local acidification and improving the quality of agricultural products and the long-term sustainability of the soil.

[0168] In summary, based on the above field tests and typical application results, it can be seen that the system described in this invention meets the technical requirements for large-scale agricultural applications in terms of power supply continuity, data stability, communication reliability, and intelligence level. It has strong environmental adaptability and engineering practicality, and is suitable for promotion and use in various agricultural scenarios such as grain crop areas, facility agriculture, high-standard farmland, and scientific research experimental areas.

[0169] IX. Summary and Description of Generalizability

[0170] In summary, this invention provides a self-powered multi-parameter farmland microenvironment monitoring system. This system integrates a multi-path micro-energy harvesting mechanism, a fusion-type multi-parameter sensing probe, an adaptive low-power wireless communication protocol, and a cloud-based intelligent modeling and decision-making module to construct an agricultural information collection platform with high reliability, low maintenance, high spatial resolution, and multi-scenario adaptability.

[0171] Compared with the prior art, the present invention achieves substantial innovation and improvement in the following aspects:

[0172] 1. In terms of energy supply mechanism, through a hybrid design of photovoltaic-thermal power-radio frequency three paths, the sensor node has achieved stable operation throughout the year in complex agricultural environment for the first time without relying on external power supply or conventional battery replacement.

[0173] 2. In terms of sensing accuracy, the integrated packaged multi-parameter probe, combined with a self-cleaning electrode structure and temperature compensation algorithm, significantly improves the long-term measurement consistency and anti-interference ability of key indicators such as soil temperature, humidity and pH.

[0174] 3. In the communication layer design, mechanisms such as adaptive adjustment of spreading factor, channel RSSI switching, and dynamic power control are introduced to systematically solve the problems of high packet loss rate and low energy efficiency of LoRa protocol under long distance and obstruction conditions;

[0175] 4. At the data application level, with the help of spatial modeling and crop water requirement model reasoning on the cloud platform, the system has the ability to perceive the environment in real time, identify anomalies and generate intelligent irrigation suggestions, forming a complete data-driven field management closed loop.

[0176] The system of this invention has been deployed and tested in multiple scenarios and across climate zones, demonstrating excellent operational stability, data reliability, and engineering maintainability in various practical application scenarios, including field crops (such as rice and wheat), facility agriculture (such as greenhouse tomatoes and cucumbers), saline-alkali land improvement, and orchard irrigation experimental fields. Its overall structure possesses good modularity and scalability, is compatible with existing agricultural IoT platforms, and has clear industrialization feasibility and commercialization prospects.

[0177] Therefore, this invention not only has significant academic research value, but also provides an intelligent sensing system solution that can be implemented in engineering and widely promoted for the digital transformation of agriculture. It has positive practical significance for promoting precision planting, green irrigation and soil health management in agriculture.

[0178] It should be understood that those skilled in the art can make various equivalent substitutions or improvements to the technical solutions of this invention without departing from the spirit and essence of this invention. All such equivalent substitutions or improvements should be considered to fall within the protection scope of this invention.

[0179] The embodiments described in this specification are only for illustrating the technical solutions of the present invention and are not intended to limit the scope of protection of the present invention. The scope of protection of the present invention should be determined by the scope defined in the appended claims.

Claims

1. A self-powered, multi-parameter farmland microenvironment monitoring system, characterized in that, include: A self-powered wireless sensor node array, each node comprising: The hybrid energy supply module consists of micro photovoltaic units and soil thermoelectric power generation units, wherein: The micro photovoltaic unit is a GaAs thin-film battery with a light conversion efficiency of not less than 28% under 200 lux conditions; The thermoelectric power generation unit includes an axially arranged Bi2Te3 thermoelectric arm, and its output power is not less than 3.2mW / cm² when the temperature difference ΔT ≥ 5K. 2 ; The multi-parameter sensing module integrates the following sensors: MEMS temperature sensor, based on Pt1000 thin film resistor, with a measurement accuracy of ±0.3℃; Interdigitated capacitive humidity sensor, measurement range 0-100%RH, error ±2%; The ISFET pH sensor uses an antimony reference electrode and has a measurement accuracy of ±0.2 pH. The communication module adopts an improved LoRa modulation method and supports dynamic switching of the spreading factor from SF7 to SF12; A gateway device, used to receive data from each node and upload it to the cloud platform, includes: An adaptive frequency hopping controller dynamically selects sub-channels in the 902-928MHz frequency band based on RSSI. The data fusion unit is used to perform kriging interpolation on spatially heterogeneous data from multiple nodes. Cloud platforms are used for data storage, visualization, and processing.

2. The system according to claim 1, characterized in that, The hybrid energy supply module also includes: The ultra-low power management chip, model BQ25570, has a startup voltage of 0.3V. The energy allocation optimization algorithm executes the following strategy:

3. The system according to claim 1, characterized in that, The hybrid power supply module also includes a radio frequency energy harvesting unit, which operates in the frequency bands of 902-928MHz and 2.4GHz, and is automatically activated when the combined output power of photovoltaic and thermoelectric power is less than 1.5mW.

4. The system according to claim 2, characterized in that, The radio frequency energy harvesting unit includes a dual-band rectifier antenna, a Schottky diode rectifier, a dynamic impedance matching network, and an energy management module, which are used to convert radio frequency signals into DC power and supply energy.

5. The system according to any one of claims 1-4, characterized in that, The system controls the operating time slot of the radio frequency rectifier through a communication scheduling mechanism to avoid communication conflicts with LoRa.

6. The system according to claim 1, characterized in that, The probe surface of the multi-parameter sensing module is provided with: The hydrophobic nano-coating has a contact angle greater than 150° and its main component is fluorinated silicon dioxide. The self-cleaning electrode structure removes contaminants by periodically applying square wave pulses ranging from -1.2V to +0.8V.

7. The system according to claim 1, characterized in that, The transmission protocol of the communication module includes: (1) The logic of the dynamic adjustment strategy of the spreading factor based on the packet bit error rate (PER) is as follows: (2) Time slot allocation method: The transmission priority is dynamically allocated based on the soil moisture value of each node, and nodes with moisture exceeding 80% are given double the transmission time slots.

8. A method for deploying a sensor network in farmland, characterized in that, include: Node deployment phase: The grid is divided according to soil type, with a deployment density of 30 nodes / hectare in clay areas and 20 nodes / hectare in sandy areas; The probe insertion depth is set to ±5cm of the main root distribution layer, with a standard of 15cm. Network self-calibration phase: The gateway issues a TDMA timeslot table, with a timeslot length of 1.3 times the data packet transmission time; The pH sensor was calibrated at two points using buffer solutions at pH=4.01 and pH=7.01, respectively.

9. A method for resisting interference from LoRa signals, characterized in that, include: Spectrum sensing steps: Detect the interference intensity in the 902-928MHz frequency band using Fast Fourier Transform (FFT); Dynamic avoidance steps: When RSSI < -90dBm and PER > 10% for a certain subchannel, switch to the backup channel; Transmit power optimization steps: Dynamically adjust the transmit power according to the following formula: P tx =min(14dBm,-0.3×distance+20) in: distance represents the distance between the node and the gateway (unit: meters); P tx This refers to the transmit power (unit: dBm). The maximum transmit power is no more than 14 dBm, and it increases linearly with distance, with a slope of -0.

3.

10. The system according to claim 1, characterized in that, The cloud platform performs the following tasks: Soil moisture prediction model: w(t+1)=0.7w(t)+0.2T(t)-0.1ETo(t) Outlier filtering algorithm: Filter pH sensor data within a range of ±3σ to remove outlier data.

Images