A near deployment boulder multi-element recoverable monitoring system and method

Abstract

This invention discloses a near-deployable multi-element retrievable monitoring system and method for root stones. The system includes a shore-based device, a sea cucumber-shaped device, a floatable tracer, and connecting pipes. The shore-based device includes measurement and control computing, an air pump, communication, and flow velocity measurement units. The sea cucumber-shaped device includes a deformation-sensing and buffering airbag and sensing and control equipment. The sensing and control equipment is used to monitor root stone attitude, vibration, deformation, and other elements for root stone stability assessment. The floatable tracer is used to activate an audible and visual alert device when it surfaces and transmits its position to the shore-based communication device. The pipes connect the shore-based device, the floatable tracer, and the sea cucumber-shaped device. This invention achieves integrated multi-element monitoring and early warning of the underwater stability of root stones, solving two major problems: instrument retrievability and avoiding damage to the instrument caused by stone throwing and compression. It also integrates sensing, positioning, information fusion, and edge computing to improve the overall system's intelligence level.

Description

Technical Field

[0001] This invention relates to a close-proximity deployment system and method for recyclable monitoring of multi-element foundation stones, belonging to the field of foundation stone detection technology for dikes. Background Technology

[0002] The stability of the foundation stones is fundamental to the safety and stability of riverbank engineering in corresponding river sections. In particular, the instability of foundation stones in dangerous sections and control sections is a major cause of engineering failures. Accurate and timely understanding of the stability status of foundation stone deformation, landslides, and vibrations is crucial for flood control and disaster relief command and decision-making. Currently, foundation stone detection in dikes commonly employs methods such as manual cone probing, mechanical cone probing, wire-type displacement gauges, tilt sensors, acoustic profilers, and underwater multibeam bathymeters. Manual methods are inherently dangerous, inefficient, and prone to misjudgment. Wire-type displacement gauges and tilt sensors have limited coverage, monitoring elements, and functions; conventionally deployed instruments do not adhere well to the foundation stones, failing to accurately measure their actual condition and are easily lost due to water flow or damaged by rock pressure. Acoustic profilers and multibeam bathymeters are not only costly and have long monitoring and analysis cycles, but are also highly susceptible to operational conditions, have numerous influencing factors affecting monitoring data, resulting in large errors, difficult and unintuitive analysis, and are challenging for long-term fixed-point monitoring. In summary, existing methods have many shortcomings in terms of long-term sustainability, effectiveness, practicality, reliability, and especially intelligence. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a near-deployable multi-element recyclable monitoring system and method for root stones. It outlines the entire process from cross-section selection, instrument installation, data analysis and early warning to instrument retrieval, achieving multi-element perception, deep fusion, edge-side computational analysis, and forecasting and early warning of root stones, thus improving the intelligence level of the monitoring system. It also saves engineering costs and avoids damage or crushing to the instrument from subsequent rockfalls and large landslides.

[0004] To achieve the above objectives, the present invention is implemented using the following technical solution:

[0005] In a first aspect, the present invention provides a near-deployable multi-element recyclable monitoring system for root stones, comprising:

[0006] The shore-based device includes a measurement and control calculation unit, an air pump, a communication unit, and a flow velocity measurement unit.

[0007] The sea cucumber-shaped device adopts a robust and waterproof bladder-like structure, carrying a sealed inflatable airbag and built-in sensing and control equipment. The sealed inflatable airbag is used to buffer the impact of the crushing rocks on the sensing and control equipment inside the robust and waterproof bladder-like structure. At the same time, the airbag has a built-in high-sensitivity gas pressure sensor to detect and monitor the deformation of the sealed airbag under the pressure of the root rocks. The built-in sensing and control equipment is used to monitor the posture, vibration and deformation of the root rocks. The data is uploaded to the onshore device for edge computing to analyze and judge the stability of the root rocks.

[0008] The floating tracer is connected to the shore device via pipes and its built-in ventilation and cables. It is used to float to the surface in case of pipe breakage, rock collapse, or the need to re-drop rocks. It can also bring the sea cucumber-shaped device to the surface and transmit its position to the shore device via wired / wireless means for easy recovery and provide sound and light alerts.

[0009] The conduit is used to protect internal cables and connect the shore-based device, the sea cucumber-like device, and the floating tracer.

[0010] Furthermore, the flow velocity measurement unit of the onshore device adopts an electromagnetic current meter or a radar current meter.

[0011] Furthermore, the floatable tracer is a sealed device with a foldable inflatable airbag. The side connected to the sea cucumber-shaped device adopts a stainless steel sealing structure. It contains / carries interconnected lithium batteries, a flashlight, a buzzer, a GPS / BD positioning device, a liquid level / hydraulic sensor, an embedded control device, and a wireless communication device. The wireless communication device can communicate with the device on shore after surfacing. The outermost side is a hard plastic shell with pre-made cracks, which is used to protect and maintain the shape of the built-in foldable airbag during installation. When floating is required, it is burst by the inflatable airbag. The stainless steel sealing structure and the plastic shell are connected by a threaded rotation, which is used to rotate a new plastic shell for reuse after recycling.

[0012] Furthermore, the pipeline is made of rubber elastic hose, with built-in power and signal cable bundles, suction cup / airbag inflation and suction pipe, high-strength recycled stainless steel wire, and pull-line displacement gauge line.

[0013] Furthermore, the sensing and control device of the sea cucumber-like device includes a sealed stainless steel body, a sensor disposed inside the sealed stainless steel body, an embedded acquisition controller, a rubber high-elasticity sealing suction cup, and an umbrella-shaped skewer assembly. The sealed stainless steel body is used to protect the internal equipment. The sensor is used to sense whether the suction cup is in contact with the intact surface of the stone, the posture of the sea cucumber-like device, vibration, and deformation between the sea cucumber-like devices. The embedded acquisition controller includes a suction cup sealing relay and an umbrella-shaped skewer assembly. The umbrella-shaped skewer assembly is used to insert into the gaps in the stone, and the rubber high-elasticity sealing suction cup is used to adhere to the surface of the root stone. Both of the above are used to fix the sea cucumber-like device to the root stone block or body, ensuring that it can monitor the stable parameters of the root stone itself.

[0014] Furthermore, the sensors within the stainless steel sealed body of the sea cucumber-shaped device include an attitude sensor, a vibration sensor, and a deformation sensor. The attitude sensor employs a high-precision electronic compass and a 3-axis tilt sensor. The electronic compass has a range of not less than 90° and an accuracy of not less than 1%FS; the 3-axis tilt sensor has a range of not less than 90° and an accuracy of not less than 0.3%FS; the vibration sensor employs a high-sensitivity triaxial accelerometer with a frequency range of 1–200Hz, a range of not less than 2g, and an accuracy of not less than 0.1%FS; the deformation sensor is a wire displacement sensor with a range of not less than 1000mm and an accuracy of not less than 0.05%FS.

[0015] Furthermore, the umbrella-shaped skewer assembly includes a root control telescopic device, an umbrella-shaped barb control and pushing device, multiple support sub-rods, a slidable rotating shaft, a fixed rotating shaft, an end with a piezoelectric sensor, and multiple umbrella-shaped barbs. The root control telescopic device is used to adjust the direction of the umbrella-shaped skewer assembly and control the telescopic movement of multiple support sub-rods within a 45-degree solid angle range (front, back, left, right) below the skewer assembly close to the root stone. The support sub-rods are connected to multiple umbrella-shaped barbs via the slidable rotating shaft. The multiple umbrella-shaped barbs are rotatably mounted on the fixed rotating shaft. By telescopically extending the support sub-rods, the multiple umbrella-shaped barbs can be opened or closed. The fixed rotating shaft is fixed in one position and can tolerate multiple support sub-rods rotating around that point at a certain angle to the central axis. The end with the piezoelectric sensor is used to locate nearby gaps. The umbrella-shaped barbs are made of a high-elasticity alloy material with an elastic diameter of 1 mm.

[0016] Secondly, the present invention provides a method for implementing a near-deployment rootstone multi-element recyclable monitoring system according to any one of the foregoing claims, comprising:

[0017] S1. Determine the monitoring section, select the dangerous section or control section with unstable foundation in history, arrange multiple measuring lines to cover the most dangerous section, and arrange the onshore device installation base and supporting civil engineering on the dike top at the corresponding chainage.

[0018] S2. Acquire underwater topographic measurement data and estimate the survey line length based on the acquired underwater topographic measurement data;

[0019] S3. Install shore-based equipment, including a shore-based communication system, an underwater blowing and suction system and suction cup attachment status judgment system, a surface flow velocity monitoring system, and an underwater sea cucumber-like device pulsation analysis software system based on surface flow velocity.

[0020] S4. Connect the sea cucumber-shaped device and the floating tracer to the onshore device through pipes, and then lay the sea cucumber-shaped device along the slope towards the foot of the root stone. When laying, proceed from the bank to the middle of the riverbed, and try to ensure that the sea cucumber-shaped device is not suspended in the air and is in contact with the root stone as much as possible.

[0021] S5. Vibration signals are collected using underwater sea cucumber-shaped devices, and the underwater vibration frequencies are compared with those obtained by inverting the measured surface flow velocity of the onshore devices to determine whether each sea cucumber-shaped device is in direct contact with the root rock or is suspended at a small height between two root rocks.

[0022] S6. If the vibration of the actual sea cucumber device is consistent with that of the calculated overhead vibration, then it is used as overhead monitoring; otherwise, it is used as adsorption monitoring.

[0023] S7. For sea cucumber-shaped devices that are determined to be non-suspended, first fix them with suction cups. If fixing with suction cups is unsuccessful, use multi-directional rods to find gaps around the device and insert them for anchoring. This will fix the sea cucumber-shaped device.

[0024] S8. Again, compare the adsorbed sea cucumber-shaped device with the suspended sea cucumber-shaped device. If there is no sea cucumber-shaped suspended device, compare it with the underwater pulsation calculation results based on the surface flow velocity of the shore to determine whether the adsorption is firm.

[0025] S9. For sea cucumber-shaped devices that are firmly attached to or inserted into the root stone, the stability of the root stone is determined by monitoring its posture, vibration, and deformation. An early warning is activated when an abnormality is detected. In the event of pipe breakage, root stone collapse, or the need for re-insertion of the stone, the airbag at the end of the buoyancy tracer is inflated according to the settings or instructions to make it float. At the same time, the suction cup of the sea cucumber-shaped device is released, the rod is retracted, and the anchoring of different sea cucumber devices to the root stone is released sequentially from the end to the shore. The device is then retrieved based on the audible and visual prompts and positioning information from the buoyancy tracer.

[0026] Furthermore, in step S6, the method for determining whether the vibration of the measured sea cucumber device is consistent with that of the calculated overhead device is as follows:

[0027] The sea cucumber-shaped device and connecting pipelines are simplified into a mass-spring-damped system. Dynamic equilibrium equations under water flow and descriptive formulas for various dynamic loads are established. The relevant mathematical model is as follows:

[0028]

[0029] In the formula: [M fs ]=ρ0[Re] T ;[K fs ] = -[Re];[Re] T =∫ s {N}{N} T {n}ds; {N} is the element shape function; {P} is the pressure vector; {U} e} represents the displacement vector; These represent the mass, damping, and stiffness matrices of the fluid, respectively; [M e [C] e ]、[K e ] are the mass, damping, and stiffness matrices of the solid, respectively; {F e} represents dynamic loads;

[0030] Since flow-induced vibration problems are generally dominated by low-order vibration modes, if we only consider q low-order vibration modes, then the above can be decomposed into q independent single-degree-of-freedom system vibration equations:

[0031]

[0032] In the formula: ω j ζ is the j-th mode circular frequency of the structure; j F is the damping ratio of the j-th mode of the structure; j (t) represents the random load component of the j-th mode of vibration of the structure, where

[0033] The transient response solution of the j-th mode can be obtained from the single-degree-of-freedom vibration equation above.

[0034] Furthermore, in step S9, the method for determining the stability of the root stone by monitoring its posture, vibration, and deformation is as follows:

[0035] In the early stages, when there was a lack of measured data, the initial values ​​of the root stone stability analysis and early warning indicators were obtained by numerical simulation using a particle flow numerical model established by a combination of computational fluid dynamics and discrete element method.

[0036] When there is a large amount of measured data in the later stages, a GRU model is established with underwater monitoring physical quantities such as root stone attitude, vibration and deformation as outputs, and multi-point water level, surface velocity and time in the river section as inputs, and the measured sea cucumber-shaped device as the monitoring data. The multi-point water level, multi-point surface velocity and time factor are inputs, and the root stone deformation, vibration and attitude measured by the sea cucumber-shaped device are outputs to establish a data-driven deep learning model. When the measured values ​​and their rate of change exceed the model prediction values, an early warning is activated.

[0037] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:

[0038] This invention provides a close-proximity, multi-element recyclable monitoring system and method for monitoring root stones. It comprehensively utilizes multiple sensors, including those for attitude, tension, vibration, and surface flow velocity, combining instrument sensing with mathematical models and integrating numerical simulation with neural network models. This achieves a high level of intelligence throughout the entire process, from installation to data analysis and early warning. Simultaneously, it enables the recovery and reuse of instruments before pipe breakage, root stone collapse, or the need for re-rocking, preventing damage to the instruments during subsequent rock removal and saving system costs. Attached Figure Description

[0039] Figure 1 This is a structural diagram of a near-deployment, multi-element recyclable monitoring system for root stones provided in an embodiment of the present invention;

[0040] Figure 2 This is a schematic diagram of the structure of the sea cucumber-shaped device provided in an embodiment of the present invention;

[0041] Figure 3 This is a schematic diagram of the structure of the buoyant tracer provided in an embodiment of the present invention;

[0042] Figure 4 This is a flowchart illustrating a method for implementing a near-deployment multi-element recyclable monitoring system for root stones, as provided in an embodiment of the present invention.

[0043] Figure 5 This is a structural diagram of the GRU unit provided in an embodiment of the present invention;

[0044] Figure 6 This is a structural diagram of the GRU neural network model provided in an embodiment of the present invention. Detailed Implementation

[0045] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0046] Example 1

[0047] like Figure 1 As shown in the figure, this embodiment introduces a near-deployed multi-element retrievable monitoring system for root stones, which consists of a shore-based device 4, a series of interconnected sea cucumber-shaped devices 2, a floating tracer 1, and a pipeline 3.

[0048] The floating tracker is a semi-open lead block with an inflatable bladder. It integrates a lithium battery, flashlight, buzzer, GPS positioning, liquid level sensor, edge computing control device, and wireless communication transmitter. It is in a dormant state underwater. When it surfaces, the audible and visual alerts activate, and it transmits its location to a shore-based communication device. During installation, the bladder folds inside the container. When recovery is needed, it inflates and floats. A vessel retrieves each tracker individually, starting from the outer end of the linear array robot, using the built-in high-strength stainless steel cable, avoiding the predicament of trackers getting stuck in gaps. The wireless communication transmitter, based on its own sensors, synchronously transmits its location information to the shore-based device and authorized mobile terminal users after surfacing.

[0049] 2- The sea cucumber-shaped device adopts an elastic, waterproof, bladder-like structure, with a built-in sealed inflatable airbag and sensing and control equipment. The airbag contains a high-precision gas pressure sensor to monitor the pressure of multiple root stones on the sealed airbag; the sensing and control equipment includes suction cups and telescopic supports, attitude sensors, vibration sensors, a self-retracting pull-wire displacement meter, an embedded edge computing unit, and telescopic supports and relay control devices. The suction cups and telescopic supports are used to attach the sea cucumber-shaped device to a solid rock or fix it in a rock crevice. The suction cups are made of corrosion-resistant underwater rubber material. One suction cup is placed directly below the turtle-back airbag, and one is placed at a 45-degree angle downwards on each side. The base of the suction cups is equipped with sensitive pressure sensors and is numbered. When the device enters the water, the relays at the base of each suction cup are in the closed state. When the pressure sensor at the base senses that the suction cup opening is pressed against the smooth wall of the rock, the corresponding relay opens, and the air pump on shore starts to pump air to ensure that the suction cup is firmly attached to the solid rock. After attachment is completed, the relays are turned off. When the sea cucumber-shaped device is between two or three rocks, multiple supports are inserted into the gaps. The heads of the supports have lantern-shaped elastic arcs to ensure that the supports are firmly stuck in the gaps.

[0050] 3-Pipeline: The flexible hose is made of highly elastic and waterproof materials such as rubber, and contains signal transmission cables, relay control and edge computing embedded processor power cables, inflation / inhalation tubes, high-strength stainless steel recycled wire, and wire for the wire displacement gauge.

[0051] 4-Onshore device: It consists of a data acquisition unit, an air pump, and a surface velocity measuring device above the riverbed root stone. The surface velocity measuring device can be implemented by an electromagnetic current meter or a radar current meter.

[0052] 5-Water surface.

[0053] like Figure 2 As shown, the sea cucumber-like device includes:

[0054] 3-1 Rubber flexible hose, used to protect internal wiring, has a certain axial elasticity and expansion performance.

[0055] 3-2 Suction cup / airbag inflation and suction tube, with a certain circumferential rigidity, spiral shape, and able to withstand a certain degree of expansion and contraction.

[0056] 3-3 Power and signal cable bundle, used to supply power to the sea cucumber-shaped device and read signals, set parameters, etc.

[0057] 3-4 high-strength recycled stainless steel wire, used to bear the force during recycling, and relaxed and not stressed during installation and operation, with a certain length reserved.

[0058] The 3-5 wire displacement gauge is in a taut state and is used to monitor the relative displacement between two adjacent sea cucumber-shaped devices.

[0059] 2-1 Sealed high-strength stainless steel body to protect internal electromechanical equipment such as chips and sensors. Includes various built-in computing and control units.

[0060] 2-2 A tortoise-shell-shaped sealing airbag protects the underlying stainless steel body and also provides shock absorption. A built-in high-precision pressure sensor detects changes in internal air pressure when the airbag is compressed by the root stone; it is a sensor for monitoring compression deformation. Its signal transmission utilizes the corresponding cables within the 3-3 signal harness.

[0061] 2-3 Sealed stainless steel internal sensors, embedded edge devices, and controllers. The sensors detect whether the suction cup is in contact with the intact surface of the rocks, the posture of the fixed sea cucumber device, and vibration. The posture sensor uses a 6-axis electronic compass and an angle sensor, while the vibration sensor uses a low-frequency, high-sensitivity triaxial accelerometer with a frequency range of 0.1–30 Hz. The controller includes a suction cup sealing relay and a fixing umbrella-shaped rod assembly, used to insert the suction cup into the gaps around the rocks like a caterpillar to fix the sea cucumber-shaped device.

[0062] 2-4 rubber high-elasticity sealing suction cups are used to adhere to the surface of the root stone and stabilize the sea cucumber-shaped device.

[0063] The 2-5 umbrella-shaped skewer assembly retracts inside the sealing device during installation. The sea cucumber-shaped device (including the turtle-shaped airbag on its back) has a total density twice that of water. When it sinks to the bottom and rests on the root stone, if the suction cup cannot adhere (judged by whether the onshore device sucks in too much water; if water is continuously sucked in, it means the suction cup cannot adhere, the water in the tube should be blown out, the corresponding solenoid valve should be closed, and the next suction cup should be replaced and tried), this umbrella-shaped skewer assembly is used. It first retracts into the gap between the root stones and then expands to stabilize. When retrieving, it first retracts and then retracts into the sealed stainless steel body.

[0064] 2-5.1 The umbrella-shaped rod group has a root control telescopic device that can adjust the direction and control the telescopic movement of the rod group within a 45-degree solid angle range;

[0065] 2-5.2 Umbrella-shaped barb control and pushing device;

[0066] 2-5.3 Support sub-rods, composed of multiple rods;

[0067] 2-5.4 Sliding hinge;

[0068] 2-5.5 Fixed pivot, fixed in one position, but can tolerate multiple support rods rotating around that point at a certain angle to the central axis, such as umbrella ribs;

[0069] 2-5.6 End with piezoelectric sensor for locating nearby gaps;

[0070] 2-5.7 Umbrella-shaped barbs, made of a high-elasticity alloy material with a diameter of 1 mm.

[0071] like Figure 3 As shown, the buoyant tracer includes:

[0072] 1-1 Control power supply and communication cable;

[0073] 1-2 GPS / BD positioning signal antennas and water pressure sensors;

[0074] 1-3 Embedded control, communication, and computing modules are used for power management, status analysis, and communication, and evaluate the buoyancy status based on pressure sensor 1-2;

[0075] 1-4 Buzzer alarm light / flashing indicator light, using sound during the day and sound and light reminders at night to facilitate location finding;

[0076] 1-5 Stainless steel sealing protective cover with O-ring sealing waterproof port, only ensuring that the vent pipe can pass through smoothly;

[0077] 1-6 Thin-walled plastic shells with pre-fabricated micro-cracks;

[0078] 1-7 Inflate the buoyancy airbag, which is in a folded state at this time to ensure that the whole body sinks to the bottom of the water. When it needs to be retrieved, it will inflate and float.

[0079] 1-8 stainless steel sealed protective cover;

[0080] 1-9 Batteries and counterweights;

[0081] 1-10 Inflation / Inhalation Tubes

[0082] Example 2

[0083] like Figure 4 As shown, this embodiment provides a method for implementing a near-deployment multi-element recyclable monitoring system for root stones according to any one of Embodiments 1, comprising:

[0084] 1. Determine the monitoring section and select the dangerous section or control section with unstable foundation in history. Generally, three measuring lines are set up to cover the most dangerous section. Onshore equipment installation base and supporting civil engineering are set up on the top of the dike at the corresponding chainage.

[0085] 2. Estimate the length of the survey line based on underwater topographic survey data, such as underwater multibeam bathymetry.

[0086] 3. Install shore-based equipment, namely shore-based communication, underwater blowing / suction and suction cup attachment discrimination system, surface flow velocity monitoring and underwater sea cucumber-like device pulsation analysis software system based on surface flow velocity.

[0087] 4. The upper end of the linear array robot is fixed and connected to the onshore device, including mechanical and electrical interfaces. Then, it is laid along the slope towards the foot of the root stone. When laying, it is carried out from the bank towards the middle of the riverbed, so as to ensure that the sea cucumber-shaped device is not suspended in the air and to make as much contact as possible with the root stone.

[0088] 5. Vibration signals were collected using underwater sea cucumber-shaped devices. The underwater vibration frequencies were compared with those obtained by inverting the measured surface flow velocity from the onshore devices to determine whether each sea cucumber-shaped device was in direct contact with the root rocks or was suspended at a small height between two root rocks.

[0089] 6. If the vibration of the actual sea cucumber device is consistent with that of the calculated overhead device, the discrimination method is to use time series similarity analysis methods such as dynamic time warping, KL divergence, and correlation analysis. If so, it is regarded as overhead monitoring; otherwise, it is regarded as adsorption monitoring.

[0090] The sea cucumber-shaped device and connecting pipelines are simplified into a mass-spring-damping system. Dynamic equilibrium equations under water flow and descriptive formulas for various dynamic loads are established. The relevant mathematical model is as follows:

[0091]

[0092] In the formula: [M fs ]=ρ0[Re] T ;[K fs ] = -[Re];[Re] T =∫ s {N}{N} T {n}ds; {N} is the element shape function; {P} is the pressure vector; {U} e} represents the displacement vector; These represent the mass, damping, and stiffness matrices of the fluid, respectively; [M e [C] e ]、[K e ] are the mass, damping, and stiffness matrices of the solid, respectively; {F e} represents the dynamic load. Since flow-induced vibration problems are generally dominated by low-order modes, if only q low-order modes are considered, the above can be decomposed into q independent single-degree-of-freedom system vibration equations:

[0093]

[0094] In the formula: ω j ζ is the j-th mode circular frequency of the structure; j F is the damping ratio of the j-th mode of the structure; j (t) represents the random load component of the j-th mode of vibration of the structure. The transient response solution of the j-th mode can be obtained from the single-degree-of-freedom vibration equation above.

[0095] 7. For sea cucumber-shaped devices that are determined to be non-suspended, first fix them with suction cups. If fixing with suction cups is unsuccessful, use multi-directional rods to find gaps around the device and insert them for anchoring. This will fix the sea cucumber-shaped device.

[0096] 8. Compare the adsorbed sea cucumber-shaped device with the suspended sea cucumber-shaped device again. If there is no suspended sea cucumber-shaped device, compare it with the underwater pulsation calculation results based on the surface flow velocity on the shore to determine whether the adsorption is firm.

[0097] 9. For firmly fixed sea cucumber-shaped devices, stability is determined by monitoring the posture and deformation of the root stones (including compression deformation and relative deformation between root stones). Initially, discrete element numerical simulation is used for this determination. Once a larger sample size is available, a GRU model is established, using underwater monitored physical quantities, multi-point water level, surface velocity, and time as inputs, and actual sea cucumber-shaped device monitoring data as outputs. Multi-point water level, multi-point surface velocity, and time are inputs, while the deformation and posture of the sea cucumber-shaped device are outputs. An early warning is activated when actual measured values ​​and their velocities exceed model predictions. An early warning is also activated immediately when multiple sensors on multiple sea cucumber-shaped devices simultaneously show abrupt changes in measurement. The inflation tube is used to inflate the end airbag, causing it to float, and the anchoring is sequentially released from the end towards the shore.

[0098] like Figure 5 As shown, the memory cells of the GRU neural network maintain two gating structures at each time step: a reset gate and an update gate. Based on these two gate structures, the GRU neural network can filter and store information.

[0099] Figure 5 Chinese x t It is the input data for the current time step after preprocessing the original upstream and downstream water levels, surface velocity, and time data of the monitoring section. t Indicates resetting the door, z t This indicates an update to the door.

[0100] (1) Reset the door

[0101] Reset door r t It is the input variable x at the current time step. t The hidden layer state h at the previous time step t-1 The data is concatenated together and transformed to the range (0,1) using the sigmoid function, representing the state h of the hidden layer at the previous time step. t-1 The percentage to be retained. The calculation for resetting the door is shown in the formula.

[0102] r t =σ(W r .[h t-1 ,x t ]+b r (1)

[0103] In the formula, W and b represent the weight matrix and bias of the gate structure, respectively, and [] denotes matrix concatenation. After calculating the reset gate, the candidate states of the hidden layer at the current time step can be obtained. The calculation is shown in the formula.

[0104]

[0105] Reset door r t The smaller the value, the less information from the hidden layer in the previous time step is retained, and the fewer candidate states of the hidden layer in the current time step. The less impact a data point receives, the better it is to reset the gate settings to uncover short-term dependencies between data points.

[0106] (2) Update Gate

[0107] Update Gate Z t Similarly, the input variable x at the current time step is... t The hidden layer state h at the previous time step t-1 The data are concatenated and transformed to the (0,1) range using the sigmoid function, but they represent different things. The update gate determines how much information from the previous time step needs to be forgotten and how much information from the current time step needs to be remembered. The calculation of the update gate is shown in the formula.

[0108] z t =σ(W z .[h t-1 ,x t ]+b z (3)

[0109] Finally, output the hidden layer state h at the current time step. t It is made by z t h t-1 , The decision is made jointly, and the calculation method is shown in the formula.

[0110]

[0111] Update Gate Z t The value of z controls whether the current state should be updated based on the state information from the previous moment. In the formula... t The closer the value of h is to 0, the better. t The closer to h t-1 h t The value of the gate is well maintained, which is key to mitigating gradient vanishing. Updating the gate value helps capture long-term dependencies between data.

[0112] From the above GRU unit structure diagram and calculation formula, it can be concluded that each GRU unit makes decisions on whether to retain or forget information, thus creating dependencies between neural network units. Typically, the hidden layers in a deep GRU neural network model are simply stacked GRU units. The deep GRU structure enables the network to have stronger non-linear learning and memory capabilities, but increasing the number of layers leads to an exponential increase in training time and memory overhead. Therefore, this paper uses a two-layer stacked GRU structure to train the network on the deformation, posture, and vibration data of the root stone measured by the sea cucumber-shaped device. The GRU neural network model structure is as follows: Figure 6 As shown.

[0113] Figure 6 x1, x2, x3…x t This represents the original impact factor input data, which is the input data of a certain data window (Window Size) after preprocessing of the upstream and downstream water levels, surface flow velocity and time data of the monitoring section. This represents the hidden state of the first-level GRU. This represents the hidden state of the second GRU layer. The Dropout layer is used to prevent overfitting by deactivating a certain percentage of neurons in the hidden layer during each iteration, ensuring the result is not overly dependent on features extracted from any particular neuron. y1, y2, y3…y t This represents the output data of the neural network, namely, the deformation, posture, and vibration of the root stone.

[0114] 10. Based on human-computer interaction and expert judgment, take measures to deal with the root stones according to the measured data. When it is determined from the measured data that it is necessary to re-insert stones, control the air inflator to inflate the end airbag, so that it floats up, and release the anchor from the end to the shore in sequence, and retrieve it.

[0115] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A near deployment boulder multi-element recoverable monitoring system, characterized by, include: The shore-based device includes a measurement and control calculation unit, an air pump, a communication unit, and a flow velocity measurement unit. The sea cucumber-shaped device adopts a robust and waterproof bladder-like structure, carrying a sealed inflatable airbag and built-in sensing and control equipment. The sealed inflatable airbag is used to buffer the impact of the crushing rocks on the sensing and control equipment inside the robust and waterproof bladder-like structure. At the same time, the airbag has a built-in high-sensitivity gas pressure sensor to detect and monitor the deformation of the sealed airbag under the pressure of the root rocks. The built-in sensing and control equipment is used to monitor the posture, vibration and deformation of the root rocks. The data is uploaded to the onshore device for edge computing to analyze and judge the stability of the root rocks. The floating tracer is connected to the shore device via pipes and its built-in ventilation and cables. It is used to float to the surface in case of pipe breakage, rock collapse, or the need to re-place rocks. It can also bring the sea cucumber-shaped device to the surface and send its own position to the shore device via wired / wireless means for easy recovery and provide sound and light alerts. The conduit is used to protect internal cables and connect the shore-based device, the sea cucumber-like device, and the floating tracer; The sensing and control device for the sea cucumber-shaped device includes a sealed stainless steel body, sensors disposed inside the sealed stainless steel body, an embedded acquisition controller, a rubber high-elasticity sealing suction cup, and an umbrella-shaped skewer assembly. The sealed stainless steel body is used to protect the internal equipment. The sensors are used to sense whether the suction cup is in contact with the intact surface of the stone, the posture of the sea cucumber device, vibration, and deformation between the sea cucumber-shaped devices. The embedded acquisition controller includes a suction cup sealing relay and an umbrella-shaped skewer assembly. The umbrella-shaped skewer assembly is used to insert into the gaps in the stone, and the rubber high-elasticity sealing suction cup is used to adhere to the surface of the root stone. Both of these are used to fix the sea cucumber-shaped device to the root stone block or body, ensuring that it can monitor the stable parameters of the root stone itself.

2. The near deployment monolith multi-element recoverable monitoring system of claim 1, wherein: The flow velocity measurement unit of the onshore device adopts an electromagnetic current meter or a radar current meter.

3. The near deployment monolith multi-element recoverable monitoring system of claim 1, wherein: The floatable tracer is a sealed device with a foldable inflatable airbag. The side connecting to the sea cucumber-shaped device uses a stainless steel sealing structure. It contains / carries an interconnected lithium battery, flashlight, buzzer, GPS / BD positioning device, liquid level / hydraulic sensor, embedded control device, and wireless communication device. The wireless communication device can communicate with the device on shore after surfacing. The outermost side is a hard plastic shell with pre-made cracks to protect and maintain the shape of the built-in foldable airbag during installation. When floating is required, the airbag will burst. The stainless steel sealing structure and the plastic shell are connected by a threaded rotation, which allows for the rotation of a new plastic shell for reuse after recovery.

4. The near deployment monolith multi-element recoverable monitoring system of claim 1, wherein: The pipeline is made of rubber elastic hose, with built-in power and signal cable bundles, suction cup / airbag inflation and suction pipe, high-strength recycled stainless steel wire, and wire displacement meter line.

5. The near-deployment multi-element recyclable monitoring system for root stones according to claim 1, characterized in that: The stainless steel sealed body of the sea cucumber-shaped device contains sensors including an attitude sensor, a vibration sensor, and a deformation sensor. The attitude sensor employs a high-precision electronic compass and a 3-axis tilt sensor. The electronic compass has a range of not less than 90° and an accuracy of not less than 1%FS; the 3-axis tilt sensor has a range of not less than 90° and an accuracy of not less than 0.3%FS. The vibration sensor employs a high-sensitivity triaxial accelerometer with a frequency range of 1–200Hz, a range of not less than 2g, and an accuracy of not less than 0.1%FS. The deformation sensor is a wire displacement sensor with a range of not less than 1000mm and an accuracy of not less than 0.05%FS.

6. The near-deployment multi-element recyclable monitoring system for root stones according to claim 1, characterized in that: The umbrella-shaped skewer assembly includes a root control telescopic device, an umbrella-shaped barb control and pushing device, multiple support sub-rods, a slidable rotating shaft, a fixed rotating shaft, an end with a piezoelectric sensor, and multiple umbrella-shaped barbs. The root control telescopic device is used to adjust the direction of the umbrella-shaped skewer assembly and control the telescopic movement of multiple support sub-rods within a 45-degree solid angle range (front, back, left, right) below the sea cucumber-shaped device close to the root stone. The support sub-rods are connected to multiple umbrella-shaped barbs via the slidable rotating shaft. The multiple umbrella-shaped barbs are rotatably mounted on the fixed rotating shaft. By telescopically extending the support sub-rods, the multiple umbrella-shaped barbs can be opened or closed. The fixed rotating shaft is fixed in one position and can tolerate the multiple support sub-rods rotating around that position at a certain angle to the central axis. The end with the piezoelectric sensor is used to locate nearby gaps. The umbrella-shaped barbs are made of a high-elasticity alloy material with an elastic diameter of 1 mm.

7. A method for implementing a near-deployment rootstone multi-element recyclable monitoring system according to any one of claims 1-6, characterized in that, include: S1. Determine the monitoring section, select the dangerous section or control section with unstable foundation in history, arrange multiple measuring lines to cover the most dangerous section, and arrange the onshore device installation base and supporting civil engineering on the dike top at the corresponding chainage. S2. Acquire underwater topographic measurement data and estimate the survey line length based on the acquired underwater topographic measurement data; S3. Install shore-based equipment, including a shore-based communication system, an underwater blowing and suction system and suction cup attachment status judgment system, a surface flow velocity monitoring system, and an underwater sea cucumber-like device pulsation analysis software system based on surface flow velocity. S4. Connect the sea cucumber-shaped device and the floating tracer to the onshore device through pipes, and then lay the sea cucumber-shaped device along the slope towards the foot of the root stone. When laying, proceed from the bank to the middle of the riverbed, and try to ensure that the sea cucumber-shaped device is not suspended in the air and is in contact with the root stone as much as possible. S5. Vibration signals are collected using underwater sea cucumber-shaped devices, and the underwater vibration frequencies are compared with those obtained by inverting the measured surface flow velocity of the onshore devices to determine whether each sea cucumber-shaped device is in direct contact with the root rock or is suspended at a small height between two root rocks. S6. If the vibration of the actual sea cucumber device is consistent with that of the calculated overhead vibration, then it is used as overhead monitoring; otherwise, it is used as adsorption monitoring. S7. For sea cucumber-shaped devices that are determined to be non-suspended, first fix them with suction cups. If the suction cups are not successfully fixed, use multi-directional rods to find the surrounding gaps and insert them for anchoring, thereby fixing the sea cucumber-shaped device. S8. Again, compare the adsorbed sea cucumber-shaped device with the suspended sea cucumber-shaped device. If there is no sea cucumber-shaped suspended device, compare it with the underwater pulsation calculation results based on the surface flow velocity of the shore to determine whether the adsorption is firm. S9. For sea cucumber-shaped devices that are firmly attached to or inserted into the root stone, the stability of the root stone is determined by monitoring its posture, vibration, and deformation. An early warning is activated when an abnormality is detected. In the event of pipe breakage, root stone collapse, or the need for re-insertion of the stone, the airbag at the end of the buoyancy tracer is inflated according to the settings or instructions to make it float. At the same time, the suction cup of the sea cucumber-shaped device is released, the rod is retracted, and the anchoring of different sea cucumber devices to the root stone is released sequentially from the end to the shore. The device is then retrieved based on the audible and visual prompts and positioning information from the buoyancy tracer.

8. The method for implementing a near-deployment multi-element recyclable monitoring system for root stones according to claim 7, characterized in that, In step S6, the method for determining whether the vibration of the measured sea cucumber device is consistent with that of the calculated overhead device is as follows: The sea cucumber-shaped device and connecting pipelines are simplified into a mass-spring-damped system. Dynamic equilibrium equations under water flow and descriptive formulas for various dynamic loads are established. The relevant mathematical model is as follows: In the formula: ; ; ; For unit shape functions; It is the displacement vector; , , These represent the mass, damping, and stiffness matrices of the fluid, respectively. , , These are the mass, damping, and stiffness matrices of the solid, respectively. For dynamic loads; Since the flow-induced vibration problem is dominated by low-order vibration modes, considering only q low-order vibration modes, the above equation can be decomposed into q independent vibration equations for a single-degree-of-freedom system: In the formula: Let j be the circular frequency of the structure's j-th mode shape; The damping ratio of the j-th mode of the structure; Let be the random load component of the j-th mode of vibration of the structure, where , The transient response solution of the j-th mode shape can be obtained from the single-degree-of-freedom vibration equation above.

Images