Multi-stage self-adaptive monitoring method and system for slope stability in typhoon storm-prone area

Abstract

The application discloses a kind of typhoon storm frequent area side slope stability multistage self-adapting monitoring method and system.Method includes: edge computing gateway gathers node electric quantity and meteorological data, and parallelly runs electric quantity loss and geotechnical mechanics coupling algorithm, calculates available energy attenuation index and dynamic disaster-causing risk index;According to risk classification, high-power consumption deformation and medium-power consumption hydrological equipment physical on-off are realized through solid-state relay, and three-level mode switching of extremely low-power consumption monitoring, hydrological perception and deformation full-dimensional monitoring is realized;At the same time, based on instantaneous rain intensity, calculate channel rain attenuation, preferentially adjust spread spectrum factor, and if necessary, increase transmission power, complete two-stage link compensation.The application constructs energy-risk double feedback system, eliminates hibernation leakage by solid-state relay isolation, combines nonlinear disaster-causing model and radio frequency reconstruction, prolongs the field survival period of equipment, and guarantees the reliable return of early warning data under heavy rain.

Description

Technical Field

[0001] This invention relates to the field of geological disaster early warning technology, specifically to a multi-level adaptive monitoring method and system for slope stability in areas prone to typhoons and rainstorms. Background Technology

[0002] Throughout the entire lifecycle of construction and operation of high-speed railways, highways, large open-pit mines, and water conservancy and hydropower projects in complex terrain conditions, the structural stability of slopes is crucial. Heavy rainfall, especially the frequent typhoons and torrential rains along the southeast coast and in the southwestern mountainous areas, is the most critical external dynamic factor inducing instability in steep rock and soil slopes. With the rapid popularization of IoT technology, deploying multi-source heterogeneous sensor networks on slope sites for unmanned, remote, online real-time monitoring has become the mainstream disaster prevention and mitigation technology standard in the engineering field.

[0003] However, existing monitoring systems have revealed fatal technical bottlenecks when dealing with extreme and prolonged weather conditions such as typhoons and torrential rains, often resulting in complete system failure on the eve of a disaster when early warnings are most needed. Specific pain points are as follows: Pain Point 1: Conventional "linear weighted early warning" and "global one-size-fits-all sampling" lead to rapid depletion of energy resources in the wild. Existing "adaptive" technologies mostly remain at a very shallow numerical logic level, relying solely on cumulative rainfall or using conventional linear weighted formulas for static assessment. Once the linearly added value exceeds the warning threshold, the system uniformly issues instructions to synchronize all high-power monitoring devices to high-frequency sampling mode. This logic violates the nonlinear mechanical process of landslide evolution in nature, leading to frequent false alarms during low-risk periods and slow response during high-risk periods when the soil is dry. Furthermore, during the long, rainy seasons of typhoons, the system prematurely depletes battery power during safe or latent periods, resulting in system failure and monitoring blind spots when disaster actually strikes.

[0004] Pain Point Two: Existing systems generally manage power consumption using software hibernation. Software hibernation involves inducing the sensor chip into a low-power mode via commands, but the power management circuitry remains physically connected to the battery, resulting in continuous microampere-level leakage current. During typhoon season, with several days or even weeks of no sunlight, this accumulated static power consumption can deplete the battery, causing the system to lose its operational capability before the disaster strikes. Furthermore, in software hibernation mode, there is a millisecond to second delay and voltage build-up process between sensor wake-up and stable operation. In scenarios requiring instantaneous capture of high-speed deformation signals preceding landslides, this wake-up delay can lead to the loss of critical data.

[0005] Pain Point 3: Ignoring the physical deprivation of underlying radio frequency links by extreme weather, resulting in "data can be collected, but cannot be transmitted." Existing systems severely neglect the significant scattering and absorption (rain attenuation effect) of microwave radio frequency signals caused by high-density water curtains and moisture during typhoons and torrential rains. Under extreme conditions where instantaneous rainfall exceeds 50 mm / h, the channel deteriorates drastically, causing a surge in packet error rates and frequent link failures between underlying nodes and edge gateways. This results in data being unable to be transmitted during critical early warning periods when slope disasters are most likely to occur and high-frequency early warning data transmission is most needed, due to the paralysis of physical communication links.

[0006] Therefore, there is an urgent need in this field for a slope intelligent monitoring solution that can deeply integrate the laws of nonlinear geotechnical mechanics, achieve precise management of the power consumption of the underlying hardware to the limit, and at the same time have the ability to adaptively adjust the radio frequency link communication. Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention provides a multi-level adaptive monitoring method and system for slope stability in areas prone to typhoons and rainstorms, aiming to solve the problems in the background technology.

[0008] To achieve the above objectives, the present invention provides the following technical solution: a multi-level adaptive monitoring method for slope stability in typhoon- and rainstorm-prone areas, comprising the following steps: Step S1: The edge computing gateway collects the power parameters of the nodes where the underlying heterogeneous monitoring devices deployed on the slope are located in real time, and at the same time obtains the meteorological data of the current environment of the slope. Step S2: Construct an energy-risk dual feedback assessment system. Based on the collected power parameters and basic meteorological data, run the power loss algorithm and the geotechnical mechanics coupling algorithm locally in parallel on the edge computing gateway to calculate the available energy decay index and dynamic disaster risk index of the current node in parallel. The available energy decay index is used to characterize the remaining survivability of the node, and the dynamic disaster risk index is used to characterize the real-time instability urgency of the slope. Step S3: Determine the slope risk status based on the comparison result between the dynamic disaster risk index and the preset threshold. The edge computing gateway performs corresponding physical on / off operations on the underlying heterogeneous monitoring equipment by controlling solid-state relays based on different risk statuses. Step S4: Calculate the physical attenuation based on the basic meteorological data of the slope, and perform two-level link adaptive compensation on the communication parameters between the edge computing gateway and the solid-state relay.

[0009] Furthermore, in step S1, the power parameters include static remaining power, dynamic concurrent load, and historical power consumption baseline; the meteorological data includes instantaneous rainfall intensity and deep soil moisture content. The underlying heterogeneous monitoring equipment is divided into three categories: high-power deformation monitoring equipment, medium-power hydrological monitoring equipment, and ultra-low-power meteorological monitoring equipment. High-power deformation monitoring equipment includes: GNSS surface displacement meters for monitoring changes in the absolute three-dimensional coordinates of the Earth's surface, and deep inclinometers for monitoring deep shear deformation zones in sliding beds. Medium-power hydrological monitoring equipment includes: shallow and deep soil moisture sensors for monitoring changes in water content during the vertical infiltration process of rainwater. Ultra-low-power meteorological monitoring equipment includes: miniature tipping bucket rain gauges for acquiring the intensity of rainstorms.

[0010] Furthermore, the specific process of step S2 is as follows: Based on the power parameters and basic meteorological data collected in step S1, the power loss algorithm and the geotechnical mechanics coupled algorithm are run locally in real time on the edge computing gateway to calculate the available energy decay index and dynamic disaster risk index of the current node in parallel. Calculate the available energy decay index using the node's static remaining power and the dynamic concurrent load of the current working cycle: ; In the formula, express The available energy decay index at any given time; express Remaining battery power at any given time point; Indicates the minimum safe sleep level; Indicates the total battery capacity at the time of manufacture; Represents the natural constant; Indicates the currently activated first Instantaneous operating power consumption of sensor-like devices; This indicates the activation duty cycle within the current acquisition period; Indicates the ambient temperature reduction factor; This indicates the total number of sensor categories that are currently active. Calculate the dynamic disaster risk index using instantaneous rainfall intensity and deep soil moisture content: ; In the formula, express Dynamic disaster risk index at any given moment; express Deep soil moisture content at any given time; This indicates the upper limit of soil saturation moisture content under the corresponding geological conditions; This represents the effective rainfall in the preceding period, adjusted for time. express Instantaneous rainfall intensity at any given moment; This represents the historical extreme value of the previous effective rainfall; This represents the historical statistical extreme values ​​of instantaneous rainfall intensity; This represents the surface displacement acceleration after noise filtering; This indicates the preset exponential amplification weighting coefficient; This represents the pre-defined overall disaster-causing weight coefficient for rainfall items; This represents the preset deformation-induced disaster weighting coefficient.

[0011] Furthermore, the underlying heterogeneous monitoring devices and edge computing gateways are connected to the main power supply, which powers them. A solid-state relay is installed between the underlying heterogeneous monitoring devices and the main power supply, and this solid-state relay is connected to the edge computing gateway. The edge computing gateway controls the opening and closing of the solid-state relay by outputting on / off commands, thereby physically switching the power supply circuit of the underlying heterogeneous monitoring devices on and off. Specifically: The high-power deformation monitoring device, the medium-power hydrological monitoring device, and the ultra-low-power meteorological monitoring device are all connected to the main power supply. A first solid-state relay is installed between the high-power deformation monitoring device and the main power supply; a second solid-state relay is installed between the medium-power hydrological monitoring device and the main power supply. The edge computing gateway integrates a first GPIO interface and a second GPIO interface. The control terminal of the first solid-state relay is connected to the first GPIO interface on the edge computing gateway via a driver circuit; the control terminal of the second solid-state relay is connected to the second GPIO interface on the edge computing gateway via a driver circuit. Both the first and second solid-state relays are normally open, and their initial state is physically open. The edge computing gateway outputs high and low level signals through the first and second GPIO interfaces to independently control the opening and closing of the first and second solid-state relays, thereby physically powering or disconnecting the high-power deformation monitoring device and the medium-power hydrological monitoring device. The ultra-low-power meteorological monitoring device is directly connected to the battery and is not controlled by any relay.

[0012] Furthermore, the underlying heterogeneous monitoring device has a built-in radio frequency chip that supports dynamic adjustment of the spread spectrum factor and transmission power, as well as an underlying node antenna. The edge computing gateway is equipped with an edge computing gateway antenna, and a wireless communication link is established between the underlying node antenna and the edge computing gateway antenna. It relies on IoT radio frequency signals to wirelessly transmit underlying monitoring data to the edge computing gateway.

[0013] Furthermore, the specific process of step S3 is as follows: Set a low-risk threshold to characterize the safety benchmark. High-risk disaster threshold characterizing the critical state of a landslide ; when Dynamic disaster risk index at any time <Low risk threshold When the slope risk status is determined to be in the risk incubation period, corresponding to the outer cloud period before the typhoon landfall, it enters the ultra-low power monitoring mode: the edge computing gateway keeps the first and second relays in the open state, or when the dynamic disaster risk index falls from the range above the low risk threshold to below the low risk threshold, the edge computing gateway issues a disconnect command to control both the first and second relays to disconnect, physically cutting off the power supply link between the high power deformation monitoring equipment and the medium power hydrological monitoring equipment; among them, the GNSS surface displacement meter only retains the ultra-low power ephemeris maintenance bypass which is not controlled by the relay and is in a warm sleep state; at the same time, only the miniature tipping bucket rain gauge is kept in a passive monitoring and interrupt triggering state. When the low risk threshold ≤ Dynamic disaster risk index at any time High-risk disaster threshold When the slope risk status is determined to be in the water infiltration softening period, corresponding to the typhoon landfall and the beginning of heavy rainfall, the hydrological sensing mode is entered: the pulse generated when the miniature tipping bucket rain gauge is tipped is used as a hardware interrupt signal to wake up the edge computing gateway. The edge computing gateway issues a closing command to control the second relay to close, while the first relay remains open. The shallow and deep soil moisture sensors begin to collect the dynamic changes in soil moisture content, providing real-time input for the moisture content in the dynamic disaster risk index, while avoiding waking up the high-power deformation monitoring equipment. when Dynamic disaster risk index at any time ≥ High-risk disaster threshold At that time, the slope risk status was determined to be in the critical period of catastrophic outbreak, corresponding to the stage where the soil is close to saturation and the catastrophic event is entering its countdown phase. The system then enters the full-dimensional deformation monitoring mode: the edge computing gateway issues a closing command, controlling the first relay connected in series with the GNSS surface displacement meter and deep inclinometer to close, thus connecting the main power supply; subsequently, the edge computing gateway retrieves the internally stored energy-frequency classification mapping table and sets the current available energy attenuation index... The numerical value is compared and matched with multiple graded threshold intervals in the energy-frequency graded mapping table, and the highest data sampling frequency allowed by the current remaining energy is issued.

[0014] Furthermore, it also includes a relay status feedback and verification mechanism: After the edge computing gateway outputs on / off commands to the corresponding first solid-state relay and second solid-state relay through the first GPIO interface and the second GPIO interface, it detects the actual voltage status of the output terminal of the corresponding solid-state relay through the analog-to-digital converter sampling circuit. If the actual status obtained by sampling is inconsistent with the expected on / off status after two consecutive on / off commands are issued, it is determined that the corresponding solid-state relay has failed, its corresponding underlying heterogeneous monitoring device node is marked as unavailable and logically isolated from the subsequent wake-up link, and a hardware fault alarm frame is generated and uploaded to the remote cloud control center through a redundant channel.

[0015] Furthermore, the specific process of step S4 is as follows: The intensity of instantaneous rainfall is quantitatively calculated using an empirical model of rainfall attenuation in the microwave band. Physical channel attenuation caused by IoT radio frequency signals transmitting early warning data between edge computing gateways and underlying heterogeneous monitoring devices : ; In the formula, express The physical attenuation of the channel at any given time; This represents the rain attenuation coefficient; Indicates the power exponent of rain decay; This represents the distance from the bottom-level node antenna to the edge computing gateway antenna; Based on the aforementioned channel physical attenuation, the dynamic underlying communication parameters, namely the spreading factor and transmit power, are solved in real time to satisfy the adaptive link compensation constraint equation: ; In the formula, for The radio frequency transmission power of the radio frequency chip in the underlying heterogeneous monitoring device is dynamically adjusted at all times. for The spreading factor is dynamically sent to the RF chip in real time. This is the receiver sensitivity threshold corresponding to the current spreading factor; For free space path loss, This represents the total gain of the transmit / receive antennas comprised of the bottom-level node antennas and the edge computing gateway antennas. For the purpose of providing a margin against multipath fading; Implement a two-stage compensation strategy, prioritizing increasing the spreading factor. To reduce the required receiver sensitivity, if the spread spectrum factor reaches the hardware limit of the RF chip and still does not satisfy the constraint equation, then the RF transmit power of the RF chip is nonlinearly increased according to the constraint equation. .

[0016] A multi-level adaptive monitoring system for slope stability in typhoon and rainstorm-prone areas includes: a multimodal heterogeneous sensor network, an edge computing gateway, and a remote cloud control center that are interconnected; the multimodal heterogeneous sensor network is distributed across the slope surface, the interior of the slope, and the safe area at the top of the slope, and the edge computing gateway is deployed at a safe high point on the slope. The multimodal heterogeneous sensor network adopts a hierarchical isolated power supply topology, specifically divided into a high-power deformation monitoring device group, a medium-power hydrological monitoring device group, and an ultra-low-power meteorological monitoring device group. The high-power deformation monitoring device group includes a GNSS surface displacement meter for monitoring changes in the absolute three-dimensional coordinates of the Earth's surface and a deep inclinometer for monitoring deep shear deformation zones in sliding beds. The medium-power hydrological monitoring device group includes soil moisture sensors arranged in a grid pattern and buried in shallow and deep soil layers for monitoring changes in water content during the vertical infiltration of rainwater. The ultra-low-power meteorological monitoring device group includes sensors for high-frequency acquisition of short-duration storm data. A miniature tipping bucket rain gauge for rainfall intensity; the graded isolated power supply topology is equipped with a battery bank, the output of which is connected to a high-power deformation monitoring device group via a first solid-state relay and to a medium-power hydrological monitoring device group via a second solid-state relay; the control terminals of the first and second solid-state relays are independently connected to the first and second GPIO interfaces of the edge computing gateway, respectively, and are subject to direct hardware on / off control by the edge computing gateway; the ultra-low-power meteorological monitoring device group is directly connected to the battery bank, is not controlled by the solid-state relays, and maintains uninterrupted operation around the clock; The edge computing gateway integrates a dual-feedback adaptive control module, a hierarchical chain wake-up control module, and a radio frequency link adaptive reconfiguration module. The dual-feedback adaptive control module is used to run the energy decay model and the dynamic disaster risk model in parallel, calculate the available energy decay index and dynamic disaster risk index of each node in real time, and output a state assessment matrix that characterizes the survivability and early warning requirements of each node. The hierarchical chain wake-up control module is used to parse the state evaluation matrix and generate high and low level control instructions through the corresponding GPIO interface according to the hierarchical logic of preset low risk threshold and high risk disaster threshold. This drives the first solid-state relay and the second solid-state relay to perform physical switching, and controls the multimodal heterogeneous sensor network to complete hardware-level switching between ultra-low power listening mode, hydrological sensing mode and deformation full-dimensional monitoring mode. The radio frequency link adaptive reconfiguration module is used to extract instantaneous rainfall intensity data from the miniature tipping bucket rain gauge as radio frequency environment input parameters, calculate the physical attenuation of the channel caused by the rainstorm to the wireless signal, and dynamically issue reconfiguration commands according to the two-level compensation strategy to adjust the spreading factor and radio frequency transmission power of the underlying wireless communication network.

[0017] Furthermore, the edge computing gateway and each sensor node of the multimodal heterogeneous sensor network aggregate data through a low-power local area network; the edge computing gateway and the remote cloud control center achieve bidirectional redundant communication through a public network communication module and a Beidou short message module to ensure the survival of the communication link when the public network is paralyzed due to extreme typhoons.

[0018] Compared with existing technologies, the present invention has the following advantages: In terms of power consumption control, this invention abandons the traditional sensor soft sleep mode. In soft sleep mode, although high-power sensor chips enter a low-power state, their power management circuits still experience continuous leakage current, resulting in significant power loss during standby periods lasting several days. This invention constructs a hierarchical chain-like hardware wake-up control mechanism. By inserting solid-state relays directly controlled by the edge computing gateway between various high-power sensors and the battery power supply, it achieves circuit-level physical power isolation for non-essential operating sensors. This hardware isolation method completely disconnects the sensors from the battery power supply circuit, eliminating static leakage current in the sleep state. Only when a disaster develops to a specific stage and the necessity of wake-up is strictly determined by the energy-risk dual feedback system, does the edge computing gateway issue hardware control signals through the GPIO interface, sequentially closing the corresponding relays to provide operating power to the corresponding sensors.

[0019] At the decision-making level, this invention constructs an energy-risk dual-feedback assessment system. This system not only calculates the dynamic disaster risk of the slope, but also introduces in parallel an energy attenuation prediction equation that includes the instantaneous power consumption of each sensor and the integral term of the activation duty cycle time, enabling the system to have a forward-looking perception of its own survivability. This decision-making mechanism allows the system to make globally optimal decisions that balance timely early warning and equipment survival rate when performing graded wake-up.

[0020] At the risk level, this invention constructs a rainfall-induced disaster coupled risk model amplified by the nonlinear exponential amplification of soil moisture content. This model sets the saturation percentage of deep soil moisture content as the nonlinear exponential amplification term of the rainfall-induced disaster driving force. When the soil is relatively dry, regardless of the rainfall amount, the overall risk index rises extremely slowly; however, when the soil moisture content approaches the physical extreme value of saturation, this exponential term exhibits an explosive growth curve, instantly amplifying the originally insignificant minute instantaneous rainfall by hundreds of times, simulating the catastrophic critical point where a sharp increase in pore water pressure leads to a precipitous drop in shear strength.

[0021] Finally, a novel approach was taken to break down the control barriers between meteorological sensing data and the underlying communication physical layer, directly inputting instantaneous rainfall intensity as a parameter into the radio frequency control closed loop of the underlying wireless communication. Utilizing empirical equations for channel physical attenuation compensation, the edge gateway proactively executes a two-stage link adaptive compensation strategy when sensing severe rainstorms: prioritizing the exchange of spreading factor for receiver sensitivity; if the spreading factor has reached its hardware limit, then nonlinearly increasing the radio frequency transmit power to counteract water curtain attenuation with electromagnetic energy. Attached Figure Description

[0022] Figure 1 This is a flowchart of the method of the present invention; Figure 2 A detailed logic flowchart of the method based on energy-risk dual feedback assessment and hierarchical chained hardware wake-up physical control provided in the embodiments of the present invention; Figure 3 This is a schematic diagram of the electrical connection logic for the hierarchical chain-type hardware wake-up control switching of three working modes in the method provided in the embodiments of the present invention. Detailed Implementation

[0023] like Figure 1 As shown, the present invention provides a technical solution: a multi-level adaptive monitoring method for slope stability in typhoon and rainstorm-prone areas. The method relies on an edge computing gateway deployed on the slope in the field and underlying multi-source heterogeneous monitoring equipment, and includes the following steps: Step S1: The edge computing gateway collects the power parameters (static remaining power, dynamic concurrent load, historical power consumption baseline) of each underlying heterogeneous monitoring device node deployed on the slope in real time, and at the same time obtains the meteorological data of the current environment of the slope (instantaneous rainfall intensity, deep soil moisture content).

[0024] Step S2: Construct an energy-risk dual feedback assessment system. Based on the electrical parameters and basic meteorological data collected in Step S1, calculate the available energy decay index and dynamic disaster risk index of the current node of the slope in parallel.

[0025] Step S3: Determine the slope risk status based on the comparison result between the dynamic disaster risk index and the preset threshold. The edge computing gateway performs corresponding physical on / off operations on the underlying heterogeneous monitoring equipment by controlling solid-state relays based on different risk statuses.

[0026] Step S4: Calculate the physical attenuation based on the basic meteorological data of the slope, and perform two-level link adaptive compensation on the communication parameters between the edge computing gateway and the solid-state relay to ensure that the early warning data of the slope can be reliably transmitted back during typhoons and rainstorms.

[0027] The underlying heterogeneous monitoring device has a built-in radio frequency chip that supports dynamic adjustment of the spread spectrum factor and transmission power, as well as an underlying node antenna. The edge computing gateway is equipped with an edge computing gateway antenna, and a wireless communication link is established between the underlying node antenna and the edge computing gateway antenna. It relies on IoT radio frequency signals to wirelessly transmit underlying monitoring data to the edge computing gateway.

[0028] In step S1, the underlying heterogeneous monitoring equipment is divided into three categories: high-power deformation monitoring equipment, medium-power hydrological monitoring equipment, and extremely low-power meteorological monitoring equipment. High-power deformation monitoring equipment includes: GNSS surface displacement gauges are used to monitor changes in the absolute three-dimensional coordinates of the Earth's surface. A deep inclinometer is used to monitor the deep shear deformation zone of a slide.

[0029] Medium-power hydrological monitoring equipment includes: Shallow and deep soil moisture sensors are used to monitor changes in water content during the vertical infiltration of rainwater.

[0030] Ultra-low power meteorological monitoring equipment includes: Miniature tipping bucket rain gauges are used to obtain short-term rainfall intensity at high frequency.

[0031] The current meteorological data is obtained through shallow and deep soil moisture sensors and miniature tipping bucket rain gauges.

[0032] The underlying heterogeneous monitoring equipment and edge computing gateway are connected to the main power supply and powered by it. A solid-state relay is installed between the underlying heterogeneous monitoring equipment and the main power supply, and this solid-state relay is connected to the edge computing gateway. The edge computing gateway controls the opening and closing of the solid-state relay by outputting on / off commands, thereby physically switching the power supply circuit of the underlying heterogeneous monitoring equipment on and off. Specifically: like Figure 2 , Figure 3As shown, the high-power deformation monitoring equipment (surface displacement gauge, deep inclinometer), the medium-power hydrological monitoring equipment (shallow and deep soil moisture sensors), and the ultra-low-power meteorological monitoring equipment (miniature tipping bucket rain gauge) are all connected to the main power supply. A first solid-state relay is installed between the high-power deformation monitoring equipment and the main power supply; a second solid-state relay is installed between the medium-power hydrological monitoring equipment and the main power supply. The edge computing gateway integrates a first GPIO interface and a second GPIO interface. The control terminal of the first solid-state relay is connected to the first GPIO interface on the edge computing gateway via a driver circuit; the control terminal of the second solid-state relay is connected to the second GPIO interface on the edge computing gateway via a driver circuit. Both the first and second solid-state relays are normally open solid-state relays, and their initial state is physically open when no control excitation signal is applied. The edge computing gateway outputs high and low level signals through the first and second GPIO interfaces to independently control the opening or closing of the first and second solid-state relays, thereby achieving physical power-off or power-on for the high-power deformation monitoring equipment and the medium-power hydrological monitoring equipment. The ultra-low power weather monitoring device (miniature tipping bucket rain gauge) is directly connected to the battery due to its microwatt-level power consumption characteristics. It is not controlled by any relays, ensuring uninterrupted 24 / 7 monitoring in any mode.

[0033] The specific process of step S2 is as follows: Based on the power parameters and basic meteorological data collected in step S1, the power loss algorithm and the geotechnical mechanics coupling algorithm are run locally in real time on the edge computing gateway to calculate the available energy decay index and dynamic disaster risk index of the current node in parallel. Calculate the available energy decay index using the node's static remaining power and the dynamic concurrent load of the current working cycle: ; In the formula, express The available energy decay index at any given time; express Remaining battery power at any given time point; This indicates the minimum power level required for safe sleep mode, which is the minimum power level required to maintain safe sleep mode for basic communication. Indicates the total battery capacity at the time of manufacture; Represents the natural constant; Indicates the currently activated first Instantaneous operating power consumption of sensor-like devices; This indicates the activation duty cycle within the current acquisition period; This represents the ambient temperature reduction factor (typically between 0.002 and 0.008 °C). This indicates the total number of sensor categories that are currently active.

[0034] Calculate the dynamic disaster risk index using instantaneous rainfall intensity and deep soil moisture content: ; In the formula, express Dynamic disaster risk index at any given moment; express Deep soil moisture content at any given time; This indicates the upper limit of soil saturation moisture content under the corresponding geological conditions; This represents the effective rainfall in the preceding period, adjusted for time. express Instantaneous rainfall intensity at any given moment; This represents the historical statistical extreme value of the previous effective rainfall, which is the historical maximum value observed in this slope area and serves as the benchmark value for normalization. The historical statistical extreme value representing the instantaneous rainfall intensity is the historical maximum value observed in this slope area, which serves as the benchmark value for normalization. It represents the surface displacement acceleration after noise filtering, which is obtained by second derivative of the three-dimensional displacement data of the GNSS surface displacement meter and then filtered, reflecting the severity and acceleration trend of slope deformation; This represents the preset exponential amplification weighting coefficient, used to control the amplification of risk by soil saturation. The larger the value, the steeper the risk growth in the saturation stage. This represents the preset overall disaster-causing weight coefficient of the rainfall item, which controls the output amplitude of "index amplification item × rainfall driving item" and matches the overall value range of the risk index. This represents the preset deformation-induced disaster weighting coefficient, used to control the contribution ratio of surface displacement acceleration to the overall risk index.

[0035] Among them, the available energy decay index The formula introduces a negative exponential decay integral term. In this integral term, Represents the currently awakened number Peak operating power consumption of sensor-like devices It's about its activation duty cycle. The core breakthrough of this algorithm lies in integrating the complex and interwoven operating frequencies of different types of sensors on the time axis into a dynamic attenuation slope of battery internal resistance consumption, thus obtaining a forward-looking dynamic sensing capability. For example, it can accurately predict the differentiated survival boundary: if the GNSS displacement meter and deep inclinometer are fully activated, the battery life will only be 2 hours; if only the rain gauge is kept on for listening, the battery life can reach 30 days, providing a quantitative basis for subsequent hierarchical decision-making. This invention constructs a nonlinear multiplicative coupling and exponential amplification equation to represent the saturation ratio of deep soil moisture content. As a nonlinear exponential amplification term of the driving force of rainfall-induced disasters. When the soil is relatively dry and the saturation is much lower than 1, regardless of the rainfall term... How large, exponent term When the soil moisture content is close to 1, the overall risk index rises extremely slowly; however, when the soil moisture content approaches its saturation physical limit, i.e. When the index approaches 1, it exhibits an exponential growth curve, instantly amplifying the originally tiny instantaneous rainfall by hundreds of times, accurately simulating the catastrophic critical point where a sharp increase in pore water pressure leads to a sudden decrease in effective stress and a precipitous drop in shear strength.

[0036] The specific process of step S3 is as follows: Set a low-risk threshold to characterize the safety benchmark. High-risk disaster threshold characterizing the critical state of a landslide Among them, the low-risk threshold High-risk disaster threshold The value is determined based on the geological parameters of the slope area and the inversion of historical rainfall-induced landslide cases. Specifically, the soil saturated moisture content is first calibrated using indoor geotechnical tests and on-site hydrological monitoring data. First, the effective rainfall reduction factor and the extreme value of instantaneous rainfall intensity are determined. Second, using a slope stability numerical model (such as the limit equilibrium method), the slope safety factor is back-calculated under different combinations of moisture content and rainfall intensity. The dynamic disaster risk index corresponding to a safety factor of 1.0 is then calibrated as the high-risk disaster threshold. The index corresponding to a safety factor of 1.5 is defined as the low-risk threshold. If detailed historical data is unavailable, A value of 0.2 to 0.3 is acceptable. A value of 0.7 to 0.8 can be used as an initial value based on engineering experience, and then dynamically corrected using on-site monitoring data.

[0037] when Dynamic disaster risk index at any time <Low risk threshold When the slope risk status is determined to be in the risk incubation period (corresponding to the outer cloud period before the typhoon landfall), it is forced to enter the ultra-low power monitoring mode: the edge computing gateway keeps the first and second relays in the open state, or when the dynamic disaster risk index falls from the range above the low risk threshold to below the low risk threshold, the edge computing gateway issues a disconnect command to control both the first and second relays to disconnect, physically cutting off the power supply link between the high power deformation monitoring equipment and the medium power hydrological monitoring equipment. Among them, the GNSS surface displacement meter only retains the ultra-low power ephemeris maintenance bypass, which is not controlled by the relay, and is in a warm sleep state (that is, through an ultra-low power circuit independent of the relay, only the ephemeris data inside the GNSS surface displacement meter is kept from being lost, so that the device is in a warm sleep state, which can save power and ensure rapid positioning within a few seconds after being woken up); at the same time, only the miniature tipping bucket rain gauge is kept in the passive monitoring and interrupt triggering state. Its function is to wake up the sleeping edge computing gateway instantly through hardware interruption, without the need for edge computing gateway polling, so that it can still respond to rainfall events with zero latency under ultra-low power.

[0038] The physical disconnection of the relay creates a physical open circuit between the power management unit of the high-power and medium-power sensors and the battery, completely eliminating the microampere-level static leakage current that is unavoidable in software sleep mode. To avoid the cold start time when the GNSS surface displacement meter is reawakened after a power outage, an extremely low-power ephemeris maintenance bypass, uncontrolled by the relay, is reserved for the GNSS surface displacement meter in a warm sleep state. In this mode, only this extremely low-power ephemeris bypass and the tipping bucket rain gauge, whose power consumption is measured in microwatts, maintain uninterrupted power supply, and the total power consumption is compressed to an extremely low level.

[0039] When the low risk threshold ≤ Dynamic disaster risk index at any time High-risk disaster threshold When the slope risk status is determined to be in the water infiltration softening period (corresponding to the typhoon landfall and the beginning of heavy rainfall), the first-level chain wake-up is triggered, entering the hydrological sensing mode: the pulse generated when the miniature tipping bucket rain gauge is turned is used as a hardware interrupt signal to wake up the edge computing gateway. The edge computing gateway issues a closing command to control the second relay to close, while the first relay remains open. The shallow and deep soil moisture sensors begin to collect dynamic changes in soil moisture content at high frequency. This can effectively provide real-time input for the amplification term of the moisture content index in the dynamic disaster risk index, while avoiding waking up high-power deformation monitoring equipment and maintaining low energy consumption.

[0040] When rainfall occurs, the miniature tipping bucket rain gauge generates a pulse signal with each tipping. These dense hardware interrupt signals wake up the corresponding GPIO interrupt service routine of the edge computing gateway via edge triggering. After being woken up, the edge gateway immediately re-runs the dual feedback evaluation and finds... The risk level has begun to cross the low-risk threshold due to increased rainfall. At this point, a strict restraint strategy is employed: instead of immediately waking up all high-power deformation monitoring devices, only one GPIO level is pulled high, which, after being driven, causes the second relay to close independently. Once the soil moisture sensor receives power, it begins high-frequency acquisition of dynamic changes in the moisture content of shallow and deep soil layers, providing crucial parameter input for subsequent calculations of the exponential amplification term. In this mode, because high-power devices such as GNSS are still physically powered off, overall power consumption remains at a low level.

[0041] when Dynamic disaster risk index at any time ≥ High-risk disaster threshold When the slope risk status is determined to be in the critical period of catastrophic outbreak (corresponding to the stage where the soil is approaching saturation and the catastrophic event is entering its countdown), the second-level chain wake-up is triggered, entering the full-dimensional deformation monitoring mode: the edge computing gateway issues a closing command, controlling the first relay connected in series with the GNSS surface displacement meter and the deep inclinometer to close, fully connecting their power supply. Since the GNSS surface displacement meter is in a pre-existing warm dormancy state maintained by ephemeris, it achieves instantaneous hot positioning without cold start delay after the main power is connected; subsequently, the edge computing gateway retrieves the internally stored energy-frequency classification mapping table and sets the current available energy attenuation index. The numerical value is compared and matched with multiple threshold intervals in the energy-frequency classification mapping table, and the highest data sampling frequency allowed by the current remaining energy is issued. The specific mapping process is as follows: when When the battery is in the high threshold range, it is determined that the power is sufficient, and the edge computing gateway issues a high-frequency sampling command (e.g., 1Hz) at the second level to obtain the most detailed deformation characteristics; when When the attenuation drops to a low threshold range, it is determined that there is a risk of power outage due to power depletion. The edge computing gateway automatically downgrades the sampling frequency to minute-level sampling (such as 0.1Hz) according to the energy-frequency classification mapping table. At the cost of sacrificing some sampling density, it ensures that the most critical disaster early warning data can obtain a longer continuous transmission cycle before the device is completely powered off.

[0042] It also includes a relay status feedback and verification mechanism: After the edge computing gateway outputs on / off commands to the corresponding first solid-state relay and second solid-state relay through the first GPIO interface and the second GPIO interface, it detects the actual voltage status of the output terminal of the corresponding solid-state relay through the analog-to-digital converter (ADC) sampling circuit. If the actual status obtained by sampling is inconsistent with the on / off status expected by the command after two consecutive on / off commands are issued, it is determined that the solid-state relay has failed, the corresponding underlying heterogeneous monitoring device node is marked as unavailable and logically isolated from the subsequent wake-up link, and a hardware fault alarm frame is generated and uploaded to the remote cloud control center through a redundant channel.

[0043] The specific process of step S4 is as follows: The intensity of instantaneous rainfall is quantitatively calculated using an empirical model of rainfall attenuation in the microwave band. Physical channel attenuation caused by IoT radio frequency signals transmitting early warning data between edge computing gateways and underlying heterogeneous monitoring devices : ; In the formula, express The physical attenuation of the channel at any given time; The rain attenuation coefficient is an empirical coefficient determined by the communication carrier frequency and antenna polarization. It characterizes the baseline amplitude of rain attenuation per unit rainfall intensity and unit transmission distance. The higher the carrier frequency, the more significant the attenuation. The larger the value, the more significant the signal attenuation effect caused by rainfall; The rain attenuation exponent is determined by the communication carrier frequency and antenna polarization, and describes the nonlinear amplification effect of rainfall intensity on signal attenuation. The closer the value is to 1, the stronger the linear correlation between rain attenuation and rainfall intensity. The larger the value, the steeper the attenuation growth in heavy rain scenarios. This represents the distance from the bottom node antenna to the edge computing gateway antenna. The longer the signal transmission path, the greater the total attenuation caused by rainfall.

[0044] Based on the aforementioned channel physical attenuation, dynamic underlying communication parameters (spreading factor and transmit power) are solved in real time to satisfy the adaptive link compensation constraint equation: ; In the formula, for The radio frequency transmission power of the radio frequency chip in the underlying heterogeneous monitoring device is dynamically adjusted at all times. for The spreading factor is dynamically sent to the RF chip in real time. This is the receiver sensitivity threshold corresponding to the current spreading factor, i.e., the minimum received power required for the gateway antenna to correctly resolve the signal; For free space path loss, This represents the total gain of the transmit / receive antennas comprised of the bottom-level node antennas and the edge computing gateway antennas. The reserved margin for multipath fading is the power redundancy reserved for slopes and outdoor terrain with multiple obstructions and reflections, which offsets the multipath signal fading caused by reflections from mountains and vegetation, and ensures link stability under extreme conditions. Implement a two-stage compensation strategy, prioritizing increasing the spreading factor. To reduce the required receiver sensitivity, if the spread spectrum factor reaches the hardware limit of the RF chip and still does not satisfy the constraint equation, then the RF transmit power of the RF chip is nonlinearly increased according to the constraint equation. .

[0045] In step S1, the precise power parameters collected by the edge computing gateway include the static remaining power of each node, the dynamic concurrent load of the current working cycle, and the historical power consumption baseline, which are used to provide a basis for calculating the available energy decay index.

[0046] The available energy decay index The negative exponential decay integral term introduced in The instantaneous power consumption of each currently activated sensor is multiplied by its activation duty cycle and summed, then integrated over time to obtain the dynamic decay slope of the battery internal resistance consumption.

[0047] Example 2

[0048] A multi-level adaptive monitoring system for slope stability in typhoon- and rainstorm-prone areas is disclosed. The system comprises a multimodal heterogeneous sensor network, an edge computing gateway, and a remote cloud control center, all interconnected. The multimodal heterogeneous sensor network is distributed across the slope surface, the interior of the slope, and the safe area at the top of the slope. The edge computing gateway is deployed at a safe high point on the slope, serving as the decision-making and control center for the entire system.

[0049] The multimodal heterogeneous sensor network adopts a hierarchical isolated power supply topology, specifically divided into a high-power deformation monitoring device group, a medium-power hydrological monitoring device group, and an ultra-low-power meteorological monitoring device group. The high-power deformation monitoring device group includes a GNSS surface displacement meter for monitoring changes in the absolute three-dimensional coordinates of the Earth's surface and a deep inclinometer for monitoring deep shear deformation zones in sliding beds. The medium-power hydrological monitoring device group includes soil moisture sensors arranged in a grid pattern and buried in shallow and deep soil layers for monitoring changes in water content during the vertical infiltration of rainwater. The ultra-low-power meteorological monitoring device group includes sensors for high-frequency acquisition of short-duration storm data. A miniature tipping bucket rain gauge for rainfall intensity; the graded isolated power supply topology is equipped with a battery bank, the output of which is connected to a high-power deformation monitoring device group via a first solid-state relay and to a medium-power hydrological monitoring device group via a second solid-state relay; the control terminals of the first and second solid-state relays are independently connected to the first and second GPIO interfaces of the edge computing gateway, respectively, and are subject to direct hardware on / off control by the edge computing gateway; the ultra-low-power meteorological monitoring device group is directly connected to the battery bank, is not controlled by the solid-state relays, and maintains uninterrupted operation around the clock.

[0050] The edge computing gateway integrates a dual-feedback adaptive control module, a hierarchical chain wake-up control module, and a radio frequency link adaptive reconfiguration module.

[0051] The dual-feedback adaptive control module is used to run the energy decay model and the dynamic disaster risk model in parallel, calculate the available energy decay index and dynamic disaster risk index of each node in real time, and output a state assessment matrix that characterizes the survivability and early warning requirements of each node.

[0052] The hierarchical chain wake-up control module is used to parse the state evaluation matrix and, based on the preset hierarchical logic of low-risk threshold and high-risk disaster threshold, generate high and low level control instructions through the corresponding GPIO interface to drive the first solid-state relay and the second solid-state relay to perform physical switching, thereby controlling the multimodal heterogeneous sensor network to complete hardware-level switching between ultra-low power monitoring mode, hydrological sensing mode, and deformation full-dimensional monitoring mode.

[0053] The radio frequency link adaptive reconfiguration module is used to extract instantaneous rainfall intensity data from the miniature tipping bucket rain gauge as radio frequency environment input parameters, calculate the physical attenuation of the channel caused by the rainstorm to the wireless signal, and dynamically issue reconfiguration commands according to the two-level compensation strategy to adjust the spreading factor and radio frequency transmission power of the underlying wireless communication network.

[0054] The edge computing gateway and each sensor node of the multimodal heterogeneous sensor network aggregate data through a low-power local area network; the edge computing gateway and the remote cloud control center achieve bidirectional redundant communication through a public network communication module and a Beidou short message module to ensure the survival of the communication link when the public network is paralyzed due to extreme typhoons.

[0055] This system fundamentally elevates the granularity of adaptive control from the software parameter adjustment level of traditional solutions to the hardware circuit on / off level, achieving the ultimate utilization of limited power in the field and ensuring the timeliness of disaster early warning and the life cycle of equipment clusters under extreme weather conditions.

[0056] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A multi-level adaptive monitoring method for slope stability in typhoon- and rainstorm-prone areas, characterized in that, Includes the following steps: Step S1: The edge computing gateway collects the power parameters of the nodes where the underlying heterogeneous monitoring devices deployed on the slope are located in real time, and at the same time obtains the meteorological data of the current environment of the slope; the underlying heterogeneous monitoring devices are divided into three categories: high power deformation monitoring devices, medium power hydrological monitoring devices, and extremely low power meteorological monitoring devices. Step S2: Construct an energy-risk dual feedback assessment system. Based on the collected power parameters and basic meteorological data, run the power loss algorithm and the geotechnical mechanics coupling algorithm locally in parallel on the edge computing gateway to calculate the available energy decay index and dynamic disaster risk index of the current node in parallel. The available energy decay index is used to characterize the remaining survivability of the node, and the dynamic disaster risk index is used to characterize the real-time instability urgency of the slope. Calculate the available energy decay index using the node's static remaining power and the dynamic concurrent load of the current working cycle: ； In the formula, express The available energy decay index at any given time; express Remaining battery power at any given time point; Indicates the minimum safe sleep level; Indicates the total battery capacity at the time of manufacture; Represents the natural constant; Indicates the currently activated first Instantaneous operating power consumption of sensor-like devices; This indicates the activation duty cycle within the current acquisition period; Indicates the ambient temperature reduction factor; This indicates the total number of sensor categories that are currently active. Calculate the dynamic disaster risk index using instantaneous rainfall intensity and deep soil moisture content: ； In the formula, express Dynamic disaster risk index at any given moment; express The deep soil moisture content at any given time; This indicates the upper limit of soil saturation moisture content under the corresponding geological conditions; This represents the effective rainfall in the preceding period, adjusted for time. express Instantaneous rainfall intensity at any given moment; This represents the historical extreme value of the previous effective rainfall; This represents the historical statistical extreme values ​​of instantaneous rainfall intensity; This represents the surface displacement acceleration after noise filtering; This indicates the preset exponential amplification weighting coefficient; This represents the pre-defined overall disaster-causing weight coefficient for rainfall items; This represents the preset deformation-induced disaster weighting coefficient; Step S3: Determine the slope risk status based on the comparison result between the dynamic disaster risk index and the preset threshold. The edge computing gateway performs corresponding physical on / off operations on the underlying heterogeneous monitoring equipment by controlling solid-state relays based on different risk statuses. The underlying heterogeneous monitoring device and edge computing gateway are connected to the main power supply and are powered by the main power supply; a solid-state relay is set between the underlying heterogeneous monitoring device and the main power supply, and the solid-state relay is connected to the edge computing gateway. Edge computing gateways control the opening and closing of solid-state relays by outputting on / off commands, thereby realizing the physical on / off of the power supply circuit of the underlying heterogeneous monitoring devices; The high-power deformation monitoring device is equipped with a first solid-state relay between itself and the main power supply; the medium-power hydrological monitoring device is equipped with a second solid-state relay between itself and the main power supply. Set a low-risk threshold to characterize the safety benchmark. High-risk disaster threshold characterizing the critical state of a landslide ; when < When the slope risk status is determined to be in the risk incubation period, the edge computing gateway enters the ultra-low power monitoring mode: the edge computing gateway keeps the first and second relays in the open state, or when the dynamic disaster risk index falls from the range above the low risk threshold to below the low risk threshold, the edge computing gateway issues a disconnect command to control both the first and second solid-state relays to disconnect. when ≤ < When the slope risk status is determined to be in the water infiltration and softening stage, the hydrological sensing mode is entered: the pulse generated when the micro tipping bucket rain gauge is tipped is used as a hardware interrupt signal to wake up the edge computing gateway. The edge computing gateway issues a closing command to control the second solid-state relay to close, while the first solid-state relay remains open. when ≥ When the slope risk status is determined to be in the critical period of disaster outbreak, it enters the full-dimensional deformation monitoring mode: the edge computing gateway issues a closing command to control the first solid-state relay to close and connect the main power supply; Subsequently, the edge computing gateway retrieves the internally stored energy-frequency classification mapping table, compares and matches the value of the current available energy decay index with multiple classification threshold ranges in the energy-frequency classification mapping table, and then issues the highest data sampling frequency allowed by the current remaining energy. Step S4: Calculate the physical attenuation based on the basic meteorological data of the slope, and perform two-level link adaptive compensation on the communication parameters between the edge computing gateway and the solid-state relay.

2. The multi-level adaptive monitoring method for slope stability in typhoon- and rainstorm-prone areas according to claim 1, characterized in that: In step S1, the power parameters include static remaining power, dynamic concurrent load, and historical power consumption baseline; the meteorological data includes instantaneous rainfall intensity and deep soil moisture content. High-power deformation monitoring equipment includes: GNSS surface displacement meters for monitoring changes in the absolute three-dimensional coordinates of the Earth's surface, and deep inclinometers for monitoring deep shear deformation zones in sliding beds; medium-power hydrological monitoring equipment includes: shallow and deep soil moisture sensors for monitoring changes in water content during the vertical infiltration of rainwater; and ultra-low-power meteorological monitoring equipment includes: miniature tipping bucket rain gauges for acquiring the intensity of rainstorms.

3. The multi-level adaptive monitoring method for slope stability in typhoon- and rainstorm-prone areas according to claim 2, characterized in that: High-power deformation monitoring equipment, medium-power hydrological monitoring equipment, and ultra-low-power meteorological monitoring equipment are all connected to the main power supply; the edge computing gateway integrates a first GPIO interface and a second GPIO interface; the control terminal of the first solid-state relay is connected to the first GPIO interface on the edge computing gateway through a driver circuit; the control terminal of the second solid-state relay is connected to the second GPIO interface on the edge computing gateway through a driver circuit; both the first and second solid-state relays are normally open solid-state relays, and their initial state is physically disconnected. The edge computing gateway outputs high and low level signals through the first and second GPIO interfaces to independently control the opening and closing of the first and second solid-state relays, thereby enabling the physical power-off or power-on of high-power deformation monitoring equipment and medium-power hydrological monitoring equipment; the ultra-low-power meteorological monitoring equipment is directly connected to the battery and is not controlled by any relay.

4. The multi-level adaptive monitoring method for slope stability in typhoon- and rainstorm-prone areas according to claim 3, characterized in that: The underlying heterogeneous monitoring device has a built-in radio frequency chip that supports dynamic adjustment of the spread spectrum factor and transmission power, as well as an underlying node antenna. The edge computing gateway is equipped with an edge computing gateway antenna, and a wireless communication link is established between the underlying node antenna and the edge computing gateway antenna. It relies on IoT radio frequency signals to wirelessly transmit underlying monitoring data to the edge computing gateway.

5. The multi-level adaptive monitoring method for slope stability in typhoon- and rainstorm-prone areas according to claim 4, characterized in that: It also includes a relay status feedback and verification mechanism: After the edge computing gateway outputs on / off commands to the corresponding first solid-state relay and second solid-state relay through the first GPIO interface and the second GPIO interface, it detects the actual voltage status of the output terminal of the corresponding solid-state relay through the analog-to-digital converter sampling circuit. If the actual status obtained by sampling is inconsistent with the on / off status expected by the command after two consecutive on / off commands are issued, it is determined that the corresponding solid-state relay has failed, its corresponding underlying heterogeneous monitoring device node is marked as unavailable and logically isolated from the subsequent wake-up link, and a hardware fault alarm frame is generated and uploaded to the remote cloud control center through a redundant channel.

6. The multi-level adaptive monitoring method for slope stability in typhoon- and rainstorm-prone areas according to claim 5, characterized in that: The specific process of step S4 is as follows: The intensity of instantaneous rainfall is quantitatively calculated using an empirical model of rainfall attenuation in the microwave band. Physical channel attenuation caused by IoT radio frequency signals transmitting early warning data between edge computing gateways and underlying heterogeneous monitoring devices : ； In the formula, express The physical attenuation of the channel at any given time; This represents the rain attenuation coefficient; Indicates the power exponent of rain decay; This represents the distance from the bottom-level node antenna to the edge computing gateway antenna; Based on the aforementioned channel physical attenuation, the dynamic underlying communication parameters, namely the spreading factor and transmit power, are solved in real time to satisfy the adaptive link compensation constraint equation: ； In the formula, for The radio frequency transmission power of the radio frequency chip in the underlying heterogeneous monitoring device is dynamically adjusted at all times. for The spreading factor is dynamically sent to the RF chip in real time. This is the receiver sensitivity threshold corresponding to the current spreading factor; For free space path loss, This represents the total gain of the transmit / receive antennas comprised of the bottom-level node antennas and the edge computing gateway antennas. For the purpose of providing a margin against multipath fading; Implement a two-stage compensation strategy, prioritizing increasing the spreading factor. To reduce the required receiver sensitivity, if the spread spectrum factor reaches the hardware limit of the RF chip and still does not satisfy the constraint equation, then the RF transmit power of the RF chip is nonlinearly increased according to the constraint equation. .

7. A multi-level adaptive monitoring system for slope stability in typhoon- and rainstorm-prone areas, used to execute the multi-level adaptive monitoring method for slope stability in typhoon- and rainstorm-prone areas as described in any one of claims 1-6, characterized in that, include: The system comprises a multimodal heterogeneous sensor network, an edge computing gateway, and a remote cloud control center that are interconnected. The multimodal heterogeneous sensor network is distributed across the slope surface, the interior of the slope, and the safe area at the top of the slope. The edge computing gateway is deployed at a safe high point on the slope. The multimodal heterogeneous sensor network adopts a hierarchical isolated power supply topology, specifically divided into a high-power deformation monitoring device group, a medium-power hydrological monitoring device group, and an ultra-low-power meteorological monitoring device group. The high-power deformation monitoring device group includes a GNSS surface displacement meter for monitoring changes in the absolute three-dimensional coordinates of the Earth's surface and a deep inclinometer for monitoring deep shear deformation zones in sliding beds. The medium-power hydrological monitoring device group includes soil moisture sensors arranged in a grid pattern and buried in shallow and deep soil layers for monitoring changes in water content during the vertical infiltration of rainwater. The ultra-low-power meteorological monitoring device group includes sensors for high-frequency acquisition of short-duration storm data. A miniature tipping bucket rain gauge for rainfall intensity; the graded isolated power supply topology is equipped with a battery bank, the output of which is connected to a high-power deformation monitoring device group via a first solid-state relay and to a medium-power hydrological monitoring device group via a second solid-state relay; the control terminals of the first and second solid-state relays are independently connected to the first and second GPIO interfaces of the edge computing gateway, respectively, and are subject to direct hardware on / off control by the edge computing gateway; the ultra-low-power meteorological monitoring device group is directly connected to the battery bank, is not controlled by the solid-state relays, and maintains uninterrupted operation around the clock; The edge computing gateway integrates a dual-feedback adaptive control module, a hierarchical chain wake-up control module, and a radio frequency link adaptive reconfiguration module. The dual-feedback adaptive control module is used to run the energy decay model and the dynamic disaster risk model in parallel, calculate the available energy decay index and dynamic disaster risk index of each node in real time, and output a state assessment matrix that characterizes the survivability and early warning requirements of each node. The hierarchical chain wake-up control module is used to parse the state evaluation matrix and generate high and low level control instructions through the corresponding GPIO interface according to the hierarchical logic of preset low risk threshold and high risk disaster threshold. This drives the first solid-state relay and the second solid-state relay to perform physical switching, and controls the multimodal heterogeneous sensor network to complete hardware-level switching between ultra-low power listening mode, hydrological sensing mode and deformation full-dimensional monitoring mode. The radio frequency link adaptive reconfiguration module is used to extract instantaneous rainfall intensity data from the miniature tipping bucket rain gauge as radio frequency environment input parameters, calculate the physical attenuation of the channel caused by the rainstorm to the wireless signal, and dynamically issue reconfiguration commands according to the two-level compensation strategy to adjust the spreading factor and radio frequency transmission power of the underlying wireless communication network.

8. The multi-level adaptive monitoring system for slope stability in typhoon- and rainstorm-prone areas according to claim 7, characterized in that: The edge computing gateway and each sensor node of the multimodal heterogeneous sensor network aggregate data through a low-power local area network; the edge computing gateway and the remote cloud control center achieve bidirectional redundant communication through a public network communication module and a Beidou short message module to ensure the survival of the communication link when the public network is paralyzed due to extreme typhoons.

Images