Ecological revetment system, self-repairing method and device for ecological revetment system

Abstract

The application provides an ecological revetment system, a self-repairing method and device of the ecological revetment system, and relates to the technical field of river revetment. The light response repair surface layer comprises a light absorption enhancement layer, a repair agent storage layer and an interface bonding layer. The piezoelectric power generation middle layer comprises a piezoelectric ceramic sheet array, and is used for generating electric energy when water flow impacts or the revetment system vibrates. The ecological concrete bottom layer comprises a porous ecological concrete layer pore structure. The edge intelligent gateway is used for dynamically adjusting the repair trigger threshold value in combination with historical crack data and sensor data fed back by a low-power sensor network, and sending a repair instruction to the ultraviolet lamp when the crack width is not less than the repair trigger threshold value, so that the ultraviolet lamp irradiates the dynamic light response material to release the repair agent to fill the cracks. The application can significantly improve the environmental adaptability of the revetment system.

Description

Technical Field

[0001] This invention relates to the technical field of riverbank protection, and in particular to an ecological bank protection system, a self-repairing method and device for the ecological bank protection system. Background Technology

[0002] Riverbank protection is a key measure to ensure riverbank stability. By reinforcing the bank slope, it can effectively prevent soil erosion and shoreline collapse, while protecting vegetation and biological habitats, promoting soil and water conservation and ecological balance. Currently, related technologies suggest the use of three types of riverbank protection structures: rigid, flexible, and ecological. However, existing riverbank protection structures mainly use solar cells to power sensors, which are prone to failure in long-term rainy or extreme environments. Furthermore, traditional riverbank protection is a monolithic cast-in-place structure, making it difficult to adapt to changes in terrain or the need for localized repairs, and exhibiting poor environmental adaptability to different riverbank protection environments. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to provide an ecological bank protection system, a self-repair method and device for the ecological bank protection system, which can significantly improve the environmental adaptability of the bank protection system.

[0004] In a first aspect, embodiments of the present invention provide an ecological revetment system, comprising: a photoresponsive repair surface layer, a piezoelectric power generation middle layer, an ecological concrete bottom layer, and an edge intelligent gateway; wherein, the photoresponsive repair surface layer comprises: a light absorption enhancement layer, a repair agent storage layer, and an interface bonding layer, and is used to soften the dynamic photoresponsive material when irradiated by natural light or ultraviolet light; the piezoelectric power generation middle layer comprises: a piezoelectric ceramic sheet array, and is used to generate electrical energy when water flow impacts or the revetment system vibrates, to drive the edge intelligent gateway and a low-power sensor network; the ecological concrete bottom layer comprises: a porous ecological concrete layer pore structure, used to place nitrifying bacteria and aquatic plant seeds in the pores for water purification and biological habitat treatment; the edge intelligent gateway is used to dynamically adjust the repair trigger threshold by combining historical crack data and sensor data fed back by the low-power sensor network, and when the crack width is detected to be not less than the repair trigger threshold, it sends a repair command to the ultraviolet light, causing the ultraviolet light to irradiate the dynamic photoresponsive material to release the repair agent to fill the crack.

[0005] In one embodiment, the low-power sensor network includes: a structural monitoring sensor, an environmental monitoring sensor, and a biological monitoring sensor; wherein the structural monitoring sensor includes: a strain sensor and a crack sensor, and is used to feed back structural data to an edge smart gateway; the environmental monitoring sensor includes: a light intensity sensor, a water level and flow rate sensor, and a water quality sensor, and is used to feed back environmental data to the edge smart gateway; the biological monitoring sensor includes: a biofilm activity sensor, and is used to feed back biological detection data to the edge smart gateway.

[0006] In one implementation, the edge smart gateway includes a lightweight AI model; wherein the lightweight AI model includes a convolutional neural network and a long short-term memory network, and the lightweight AI model is used to determine the repair trigger threshold and repair mode based on historical crack data and sensor data.

[0007] Secondly, embodiments of the present invention also provide a self-repair method for an ecological revetment system. The method is applied to an edge smart gateway of the ecological revetment system. The method includes: acquiring historical crack data, structural data, environmental data, and biological detection data from the ecological revetment system, as well as energy data from the piezoelectric power generation midlayer; and determining input data with a unified data structure by aligning the data according to timestamps; performing feature extraction processing on the input data to determine the model input vector, and sending the model input vector to a lightweight AI model for dynamic threshold prediction and mode selection processing to determine the repair trigger threshold and repair mode; when the crack width collected by the crack sensor is detected to be not less than the repair trigger threshold, sending a repair command to the ultraviolet lamp through a low-power wide area network communication protocol, causing the ultraviolet lamp to irradiate the dynamic photoresponsive material to release a repair agent to repair the crack.

[0008] In one implementation, after determining the input data with a uniform data structure, the method includes: denoising the input data using a sliding window averaging denoising method, and performing temperature drift calibration on the denoised data to determine the target input data.

[0009] In one implementation, the step of performing feature extraction processing on the input data to determine the model input vector includes: performing feature extraction processing on the crack propagation rate, illumination trend and water flow impact frequency in the target input data to determine the feature vector; performing standardization processing on the feature vector using a preset Z-Score normalization processing model and a preset maximum-minimum scaling model; and performing data fusion processing on the standardized feature vector to determine the model input vector.

[0010] In one implementation, the steps of sending the model input vector to a lightweight AI model for dynamic threshold prediction and mode selection processing to determine the repair trigger threshold and repair mode include: using a convolutional neural network and a long short-term memory network in the lightweight AI model to perform dynamic threshold prediction processing on the model input vector to determine the repair trigger threshold; performing mode selection processing based on the energy storage capacity in the energy data and the water flow velocity in the environmental data; when the energy storage capacity and water flow velocity are within the preset parameter threshold range, the repair mode is determined to be passive repair, and natural light is used to initiate crack repair; when the energy storage capacity and water flow velocity are not within the preset parameter threshold range, the repair mode is determined to be active repair, and ultraviolet lamps are used to initiate crack repair.

[0011] Thirdly, embodiments of the present invention also provide a self-repair device for an ecological revetment system. The device is applied to the edge smart gateway of the ecological revetment system and includes: a data acquisition module, which acquires historical crack data, structural data, environmental data, and biological detection data from the ecological revetment system, as well as energy data from the piezoelectric power generation midlayer, and determines input data with a unified data structure by aligning the data according to timestamps; a data analysis module, which performs feature extraction processing on the input data, determines the model input vector, and sends the model input vector to a lightweight AI model for dynamic threshold prediction and mode selection processing to determine the repair trigger threshold and repair mode; and a repair control module, which, when the crack width collected by the crack sensor is not less than the repair trigger threshold, sends a repair command to the ultraviolet lamp through a low-power wide area network communication protocol, causing the ultraviolet lamp to irradiate the dynamic photoresponsive material to release a repair agent to repair the crack.

[0012] Fourthly, embodiments of the present invention also provide a server, including a processor and a memory, the memory storing computer-executable instructions executable by the processor, the processor executing the computer-executable instructions to implement any of the methods provided in the first aspect.

[0013] Fifthly, embodiments of the present invention also provide a computer-readable storage medium storing computer-executable instructions, which, when invoked and executed by a processor, cause the processor to implement any of the methods provided in the first aspect.

[0014] The embodiments of the present invention bring the following beneficial effects:

[0015] This invention provides an ecological revetment system, a self-repair method for the ecological revetment system, and an apparatus. The system includes: a photoresponsive repair surface layer, a piezoelectric power generation middle layer, an ecological concrete bottom layer, and an edge smart gateway. The photoresponsive repair surface layer includes: a light absorption enhancement layer, a repair agent storage layer, and an interface bonding layer. This layer softens the dynamic photoresponsive material under natural light or ultraviolet light irradiation. The piezoelectric power generation middle layer includes: a piezoelectric ceramic sheet array. This layer generates electricity when water flow impacts or the revetment system vibrates, driving the edge smart gateway and a low-power sensor network. The ecological concrete bottom layer includes: a porous ecological concrete layer with a porous structure for storing nitrifying bacteria and aquatic plant seeds within the pores. This invention aims to purify water and treat biological habitats. An edge smart gateway combines historical crack data with sensor data from a low-power sensor network to dynamically adjust the repair trigger threshold. When the crack width is detected to be no less than the repair trigger threshold, a repair command is sent to the ultraviolet lamp, causing the ultraviolet lamp to irradiate the dynamic photoresponsive material to release a repair agent that fills the crack. This embodiment of the invention can repair the surface layer based on photoresponsiveness, enabling ultraviolet light to trigger the rapid release of the repair agent to fill the crack, breaking through the limitations of traditional temperature and salinity environments, significantly improving repair efficiency and environmental adaptability. Furthermore, by integrating a lightweight AI model, it can analyze multi-source data in real time and dynamically adjust the repair threshold, optimizing repair timing and mode, reducing cloud dependence, and improving real-time response and energy efficiency.

[0016] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained in accordance with the structures particularly pointed out in the description, claims and drawings.

[0017] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0018] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of the structure of a self-repairing system for an ecological bank protection system provided in an embodiment of the present invention;

[0020] Figure 2 This is a schematic diagram of the structure of a self-repairing system for another ecological bank protection system provided in an embodiment of the present invention;

[0021] Figure 3 A flowchart illustrating a self-repair method for an ecological bank protection system provided in an embodiment of the present invention;

[0022] Figure 4 This is a schematic diagram of the structure of a self-repairing device for an ecological bank protection system provided in an embodiment of the present invention;

[0023] Figure 5 This is a schematic diagram of the structure of a server provided in an embodiment of the present invention. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] Currently, bank protection is a key measure to ensure riverbank stability. By reinforcing the bank slope, it effectively prevents soil erosion and shoreline collapse, avoids damage to farmland and buildings, regulates water flow patterns, reduces the risk of flood erosion, and maintains the safety of river flood control. At the same time, ecological bank protection takes into account vegetation restoration and habitat protection, promotes soil and water conservation and ecological balance, and plays an important role in disaster prevention and mitigation, protecting the lives and property of riverside residents and sustainable development. Related technologies suggest that existing bank protection is mainly divided into rigid (concrete, masonry, etc.), flexible (gabion, eco-bag, etc.), and ecological (vegetated slope protection, fish nest bricks, etc.). Rigid structures have strong erosion resistance, ecological structures take into account both protection and natural restoration, and flexible structures combine stability and permeability to adapt to different river channel needs.

[0026] However, existing bank protection technologies still have the following shortcomings: (1) Limited self-repair efficiency: Existing microbial-induced repair relies on specific environments (such as pH and temperature), and the repair effect is significantly reduced in low-temperature or high-salinity waters; (2) Monitoring system relies on external power supply: Existing sensor networks rely on solar energy and batteries, which are prone to failure in long-term rainy or extreme environments; (3) Single ecological function: Bank protection structures only focus on protection and repair, lacking collaborative design for water purification and biodiversity enhancement; (4) Poor construction flexibility: Traditional bank protection is integrally cast, which is difficult to adapt to changes in terrain or local repair needs. Based on this, the ecological bank protection system, the self-repair method and device of the ecological bank protection system provided by this invention can repair the surface based on light response, so that ultraviolet light triggers the rapid release of repair agents to fill cracks, breaking through the traditional temperature and salinity environment limitations, significantly improving repair efficiency and environmental adaptability, and by integrating a lightweight AI model, analyzing multi-source data in real time and dynamically adjusting the repair threshold, optimizing the repair timing and mode, reducing cloud dependence, and improving the real-time response and energy efficiency ratio.

[0027] To facilitate understanding of this embodiment, a self-repair method for an ecological revetment system disclosed in this invention will first be described in detail. This method is applied to an ecological revetment system. To facilitate understanding of the ecological revetment system, this invention provides a structural schematic diagram of an ecological revetment system, as shown below. Figure 1 As shown, the system includes: a photoresponsive self-healing surface layer, a piezoelectric power generation middle layer, an ecological concrete bottom layer, and an edge smart gateway; among which, the modular revetment unit composed of the photoresponsive self-healing surface layer, the piezoelectric power generation middle layer, and the ecological concrete bottom layer has a planar dimension of 1m × 1m and a thickness of 10~20cm. Additionally, see [link to documentation]. Figure 2 The diagram shows the structure of another ecological revetment system with a self-repair system. The modules are connected by mortise and tenon joints. The top of the revetment unit is capped with concrete to connect with the existing terrain, and the bottom is supported and reinforced with a chamfered concrete foundation. The modular revetment unit is embedded with a through-pore structure (pore diameter 0.1-5mm), which contains nitrifying bacteria and aquatic plant seeds, thus simultaneously achieving water purification and biological habitat functions.

[0028] The photo-responsive repair surface layer includes: a light absorption enhancement layer, a repair agent storage layer, and an interface bonding layer. The photo-responsive repair surface layer is used to soften the dynamic photo-responsive material when exposed to natural light or ultraviolet light. The photo-responsive self-repairing surface layer uses a photosensitive polymer as the matrix and encapsulates the repair agent (nano silica and plant-extracted adhesive). Under ultraviolet light, it triggers the material to soften and release the repair agent to fill the crack. Its material composition is: photosensitive polymer (polycaprolactone-PCL, accounting for 40%), nano silica (particle size 20~50nm, 20%), plant-extracted adhesive (such as natural polysaccharides, 20%), toughening fiber (length 1-3mm, carbon fiber, 10%), photoinitiator (benzophenone, 10%), and the auxiliary agent is mainly a plasticizer (glycerin, accounting for 5%), which is used to improve the flexibility of the material. Specifically, it includes the following (1) to (4):

[0029] (1) Pretreatment and dispersion: 1. Modification of nano silica: Place nano silica in an ethanol solution of silane coupling agent (such as KH-550) and sonicate for 30 minutes to improve its compatibility with the polymer matrix; 2. Activation of plant adhesive: Dissolve the plant-extracted adhesive in deionized water (concentration 20%) and stir at 60°C until completely dissolved; 3. Surface treatment of carbon fiber: Oxidize the surface of carbon fiber with nitric acid to enhance its interfacial bonding with the matrix material.

[0030] (2) Mixing and blending: 1. Melt blending: Heat PCL particles to 160°C to melt, then add modified nano silica, activated plant binder solution, carbon fiber and benzophenone in sequence, and keep the temperature constant while stirring (200 rpm, 30 minutes); 2. Plasticizer addition: Add glycerol and continue stirring for 10 minutes to ensure the system is homogenized; 3. Vacuum degassing: Transfer the mixture to a vacuum reactor and evacuate to -0.1 MPa to remove bubbles for 20 minutes.

[0031] (3) Molding and curing: 1. Mold molding: Inject the degassed melt into a pre-made mold (thickness 1-2mm) and cool to room temperature for curing; 2. UV pre-activation: Expose the molding material to UV light for a short time (wavelength 365nm, intensity 50mW / cm², time 5 minutes) to initiate partial cross-linking of the photoinitiator and initially form a photoresponsive network; 3. Post-curing treatment: Place in a 50℃ oven and let stand for 24 hours to eliminate internal stress and stabilize the material structure.

[0032] (4) Performance verification and optimization: 1. Repair efficiency test: A 2mm crack was artificially created and irradiated with a UV lamp (intensity 200W / m²) for 10 minutes. The crack closure rate and the formation of the silicone network were observed. 2. Mechanical performance test: The elongation at break (≥200%) and compressive strength (≥5MPa) of the material were measured by a tensile testing machine. 3. Environmental adaptability verification: The release rate of the repair agent and the durability of the material were tested by simulating low temperature (-20℃) and high salt (3% NaCl solution) environments.

[0033] In a real-time approach, after crack formation, ultraviolet (UV) irradiation softens the PCL, releasing nano-silica and an adhesive. The adhesive expands upon contact with water, combining with the silica to form a silicone network that rapidly fills the crack. The thickness parameter can be determined according to the Beer-Lambert law, and the material thickness d must satisfy a UV transmittance T ≥ 50%, calculated using the following formula:

[0034]

[0035] in, The attenuation coefficient of the photosensitive polymer can be taken as 0.7~0.8 mm⁻¹.

[0036] Therefore, material thickness mm, which can generally be taken as 0.8 mm.

[0037] The photoresponsive structure employs a gradient composite structure, comprising a surface layer (light absorption enhancement layer), a middle layer (repair agent storage layer), and a bottom layer (interfacial bonding layer). The surface layer (light absorption enhancement layer) is 0.1 mm thick and contains a photosensitive polymer doped with titanium dioxide nanoparticles (5 wt%) to enhance UV absorption efficiency and concentrate UV energy on the surface, accelerating material softening. The middle layer (repair agent storage layer) is 0.6 mm thick and contains repair agent microcapsules (nano-silica + plant-based binder) with a diameter of 50–100 μm, accounting for 30% of the volume. A composite of the photosensitive polymer and carbon fiber (10 wt%) serves as the supporting matrix, enhancing crack resistance. The microcapsules are arranged in a hexagonal close-packed configuration with a density of 1.2 × 10⁴ capsules / cm², ensuring uniform distribution of the repair agent. The bottom layer (interfacial bonding layer) is 0.1 mm thick and primarily consists of an epoxy resin adhesive layer (containing a silane coupling agent) to enhance the bonding strength with the piezoelectric power generation layer.

[0038] The bottom layer of the ecological concrete includes: a porous ecological concrete layer with a porous structure, which is used to set nitrifying bacteria and aquatic plant seeds in the pores for water purification and biological habitat treatment.

[0039] The edge smart gateway combines historical crack data and sensor data from a low-power sensor network to dynamically adjust the repair trigger threshold. When the crack width is detected to be no less than the repair trigger threshold, it sends a repair command to the ultraviolet lamp, causing the lamp to irradiate a dynamically photoresponsive material to release a repair agent that fills the crack. The edge smart gateway includes a lightweight AI model, which comprises a convolutional neural network and a long short-term memory network. This lightweight AI model determines the repair trigger threshold and repair mode based on historical crack data and sensor data. The edge smart gateway uses an STM32F7 series microcontroller (216MHz clock speed, ≥1MB memory) and integrates a LoRaWAN communication module. One gateway is configured for every 50 revetment modules, fixed to the top of the revetment or adjacent structures.

[0040] The low-power sensor network includes structural monitoring sensors, environmental monitoring sensors, and biological monitoring sensors. The network can utilize ultra-low-power LoRaWAN sensors to monitor strain, cracks, water quality (COD, pH), and biofilm activity. The structural monitoring sensors include strain sensors and crack sensors. These sensors feed structural data back to the edge smart gateway. Specifically, strain sensors are located at the four corners and center of each revetment unit (5 per module) to monitor unit deformation, stress distribution, and identify localized overload areas. Crack sensors are linearly arranged along the module joints at 20cm intervals (2 sensors per module with 4 joints) to detect crack width and propagation rate in real time, prioritizing coverage of mortise and tenon joints.

[0041] The environmental monitoring sensors include: a light intensity sensor, a water level and flow velocity sensor, and a water quality sensor. These sensors feed environmental data back to the edge smart gateway. Specifically, the light intensity sensor is located in the light-responsive material area on the revetment surface, with one sensor at the top center of each module, used to monitor ultraviolet radiation intensity and trigger the self-healing material response. The water quality sensors (COD, pH) are located at the bottom of the porous biological carrier layer, with two sensors evenly distributed per module (near the water inlet and outlet), used to assess the water purification effect and provide feedback on ecological restoration efficiency. The water level and flow velocity sensors are located on the water-facing side of the revetment, arranged at 10m intervals, used to monitor hydrological dynamics and optimize piezoelectric power generation efficiency.

[0042] The bio-monitoring sensors include: a biofilm activity sensor, which is used to feed back bio-detection data to the edge smart gateway. Specifically, the biofilm activity sensor is located in the ecological concrete layer (microbial enrichment zone), with one sensor arranged in each module to detect the metabolic activity of nitrifying bacteria and evaluate the ecological synergy effect.

[0043] In other words, a single module corresponds to 17 sensors, including: structural monitoring: strain sensors (5) + crack sensors (8), totaling 13 per module; environmental monitoring: light (1) + water quality (2), totaling 3 per module; biological monitoring: biofilm activity sensor (1), 1 per module.

[0044] The piezoelectric power generation middle layer includes a piezoelectric ceramic sheet array. This middle layer generates electricity when impacted by water flow or when the bank protection system vibrates, driving edge smart gateways and low-power sensor networks. A single module generates ≥5Wh per day, with a total thickness of approximately 3mm. It includes piezoelectric sheets and an encapsulation layer. The encapsulation layer further comprises an upper protective layer and a lower support layer. The piezoelectric sheets are PZT-5H ceramic sheets with a single-layer thickness of 1.5~2.0 mm (balancing power generation efficiency and mechanical strength). Experiments have verified that a 1.0 mm thick PZT-5H sheet achieves a power density of 0.8 mW / cm² under a 2 m / s water flow. The upper protective layer of the encapsulation layer is 0.5~0.6 mm thick and is a silicone layer for waterproofing and impact resistance. The lower support layer is 1.5~2.0 mm thick and is made of glass fiber reinforced epoxy resin to improve bending stiffness.

[0045] In summary, the minimum thickness of the piezoelectric power generation layer is 8mm. According to the power generation efficiency model, the relationship between the piezoelectric output power P and the thickness t is as follows:

[0046]

[0047] in, It is the piezoelectric constant. Let f be the dielectric constant, f be the vibration frequency, and A be the area.

[0048] The piezoelectric elements are arranged in a diamond grid (5 mm spacing). Each module (1m×1m) contains 400 PZT-5H elements (50mm×20mm×1.0mm in size). The circuit is connected in parallel, which can reduce impedance and increase current output (peak current of a single module ≥20 mA).

[0049] The piezoelectric layer and the photoresponsive material layer are bonded together with epoxy adhesive (peel strength ≥30 N / cm) and connected to the ecological concrete layer.

[0050] In one implementation, for the distribution of piezoelectric power generation energy, a dual supercapacitor parallel connection (10F, 5.5V per capacitor) energy storage design can be adopted to achieve redundant power supply (maintaining power supply capability when a single capacitor fails). The priority of its intelligent power distribution strategy is set so that the power supply priority of the crack sensor and the gateway is the highest, followed by the water quality sensor. It can also be dynamically adjusted according to the energy storage capacity (when the remaining capacity is <20%, only power is supplied to the critical nodes). With dual supercapacitor parallel energy storage, 70% power supply capability is maintained when a single capacitor fails, and power is supplied only to the critical sensor nodes when the remaining capacity is <20%.

[0051] In another implementation, a low-power mode design is also adopted. The sensor sleep mechanism switches to sleep mode during inactive periods (such as nighttime without light), reducing power consumption to 0.1mW. The data acquisition cycle is a default wake-up every 10 minutes. In emergency situations (such as crack propagation rate > 0.1mm / h), it switches to real-time monitoring.

[0052] The ecological revetment system provided in this invention adopts modular revetment units connected by mortise and tenon joints, allowing for rapid assembly and adaptation to complex terrain. In case of localized damage, only a single module needs to be replaced, significantly reducing construction and maintenance costs. The revetment surface uses a photoresponsive self-healing material composed of a photosensitive polymer (polycaprolactone-PCL), nano-silica, plant-derived adhesive, and toughening fibers. Under ultraviolet irradiation, the material softens and releases a repair agent, filling cracks. The system's upper layer integrates a piezoelectric power generation device, utilizing water flow fluctuations and revetment vibrations to generate electricity, powering a low-power sensor network and edge computing gateway, achieving energy self-sufficiency. The porous structure of the porous ecological concrete layer, embedded with nitrifying bacteria and aquatic plant seeds, simultaneously achieves water purification and biological habitat functions. The edge intelligent gateway incorporates a lightweight AI model, dynamically adjusting the repair threshold based on historical data and real-time environmental parameters to optimize the repair triggering timing. This system boasts advantages such as strong environmental adaptability, zero external power supply, and integrated ecological functions. It is suitable for various aquatic environments, and exhibits excellent stability and practicality, especially under extreme climatic conditions, providing an innovative solution for the field of river ecological bank protection.

[0053] based on Figure 1 The diagram shows the structure of a self-repairing system for an ecological bank protection system. Figure 2 The diagram shown illustrates the structure of another self-repairing ecological revetment system. This invention provides a detailed description of the self-repairing method for ecological revetments. (See attached diagram.) Figure 3 The diagram shows a self-repair method for an ecological bank protection system, which mainly includes the following steps S302 to S306:

[0054] Step S302: Acquire historical crack data, structural data, environmental data, and biological detection data from the ecological revetment system, as well as energy data from the piezoelectric power generation layer. Align the data by timestamp to determine the input data for a unified data structure. The data types include: Structural data (strain sensors, deformation monitoring), crack sensors (crack width, propagation speed); Environmental data (light intensity sensors, ultraviolet intensity), water level sensors, water quality sensors (COD, pH), and water flow velocity sensors); Energy data (piezoelectric layer power generation, remaining battery power). In one embodiment, the sensor data are aligned by timestamp and converted into a unified data structure.

[0055]

[0056] Where t is the timestamp and x is the value of each sensor. In response, For cracks, For light, For water flow, It represents voltage.

[0057] In one implementation, the input data also needs to be denoised using a sliding window averaging denoising method, and the denoised data needs to be temperature drift calibrated to determine the target input data. Specifically, the sliding window averaging denoising method is used, with a window length N=10 (corresponding to a 1s frequency and a 10Hz sampling rate) and a sliding step size S=1. The denoised data is then:

[0058]

[0059] For example, the original data sequence of the strain sensor is , , ..., The data for the 10th data point after noise reduction is as follows:

[0060]

[0061] Temperature drift calibration. Establish a linear relationship between temperature T and sensor reading x:

[0062]

[0063] The coefficients a and b were determined through calibration experiments (sensor reference values ​​were collected at different temperatures in a constant temperature chamber, and a straight line was fitted).

[0064] Step S304: Perform feature extraction processing on the input data to determine the model input vector, and send the model input vector to the lightweight AI model for dynamic threshold prediction and mode selection processing to determine the repair trigger threshold and repair mode. In one embodiment, this invention also provides an implementation method for an automatic ecological revetment system, as detailed in (A) to (C) below:

[0065] (A) The crack propagation rate, illumination trend and water flow impact frequency in the target input data are processed by feature extraction to determine the feature vector. The feature vector is then standardized using the preset Z-Score normalization model and the preset maximum-minimum scaling model. The standardized feature vector is then processed by data fusion to determine the model input vector, specifically including the following (1) to (5):

[0066] (1) Crack propagation rate. For crack width... Perform a first-order difference to calculate the instantaneous spread rate:

[0067]

[0068] right Applying the exponentially weighted moving average, we obtain:

[0069]

[0070] (2) Light intensity trend analysis. The average light intensity over the past 5 minutes was calculated using the moving average method:

[0071] (Sampling interval 1s)

[0072] By fitting a linear trend using the least squares method, we can obtain:

[0073]

[0074] (3) Water flow impact frequency. Fourier transform is used to process the water flow velocity signal. The extracted main frequency component is:

[0075]

[0076] The frequency with the largest amplitude in the frequency spectrum is: .

[0077] (4) Data standardization

[0078] The Z-Score normalization method is used to scale each feature to a distribution with a mean of 0 and a standard deviation of 1:

[0079]

[0080] in, This is the average of historical data; The standard deviation is denoted as .

[0081] And perform maximum-minimum scaling on environmental parameters such as light intensity:

[0082]

[0083] (5) Data fusion and output: The preprocessed data is integrated into the model input vector.

[0084]

[0085] (B) Utilizing the convolutional neural network and long short-term memory network in the lightweight AI model, dynamic threshold prediction processing is performed on the model input vector to determine the repair trigger threshold. In one implementation, the lightweight AI model adopts a fusion architecture of convolutional neural network (CNN) and long short-term memory network (LSTM), and achieves efficient inference through model pruning, quantization, and hardware acceleration techniques. The input layer receives the preprocessed feature vector. The lightweight CNN branch and the lightweight LSTM branch are constructed, specifically including the following (1) to (3):

[0086] (1) Lightweight CNN branch. Depthwise separable convolution is used instead of standard convolution, which reduces the computation to 1 / N (N is the number of input channels).

[0087]

[0088] in, Where M is the convolution size and M is the number of output channels. The size of the feature map.

[0089] (2) Lightweight LSTM branch

[0090] The LSTM hidden layers are divided into 4 groups, and each group is calculated independently to reduce the number of parameters.

[0091]

[0092] in, For input dimensions, This represents the size of the hidden layer.

[0093] (3) Branch fusion

[0094] The outputs of the CNN and LSTM are concatenated, and dimensionality is reduced by 1×1 convolution.

[0095] In one implementation, the CNN branch extracts spatial patterns by passing the input features through three depthwise separable convolutions (3×3 kernel size, 16-32-64 channels); the LSTM branch passes the input temporal data through a grouped LSTM (64-16*4 hidden layers) to output temporal features; and the feature fusion concatenates the outputs of the CNN and LSTM, reduces the dimensionality through a 1×1 convolution (128-64 channels), and then inputs it into a fully connected layer.

[0096] Based on the aforementioned historical crack data (maximum width and frequency of occurrence within 72 hours), real-time light intensity, water flow velocity, and energy storage capacity, if the light intensity remains below 200W / m² (during cloudy or rainy conditions), the fully connected layer will output a dynamic threshold. :

[0097]

[0098] in, For example, the threshold is reduced from the default 0.5mm to 0.3mm, and when the crack reaches 0.3mm, repair is triggered in advance.

[0099] (C) Based on the energy storage capacity in the energy data and the water flow velocity in the environmental data, a mode selection process is performed. When both the energy storage capacity and the water flow velocity are within the preset parameter threshold range, the repair mode is determined to be passive repair, using natural light to open the crack for repair. When neither the energy storage capacity nor the water flow velocity is within the preset parameter threshold range, the repair mode is determined to be active repair, using ultraviolet lamps to open the crack for repair. In a real-time mode, the repair mode is selected based on the energy storage capacity E and the water flow velocity:

[0100]

[0101] Therefore, when the energy storage capacity is ≥50% and the flow rate is ≤2m / s, the crack is repaired through the module's "passive repair" (relying on natural light) mode; under other conditions, the crack is repaired by activating "active repair" (calling the backup power to start the UV lamp).

[0102] Step S306: When the crack width detected by the crack sensor is not less than the repair trigger threshold, a repair command is sent to the ultraviolet lamp via the LoRaWAN communication protocol, causing the ultraviolet lamp to irradiate the dynamic photoresponsive material to release the repair agent and repair the crack. In one embodiment, when the crack width is greater than or equal to the dynamic threshold, the gateway generates a repair command, which includes the following parameters: 1. Repair range: locate the damaged module number according to the crack distribution; 2. Repair intensity: graded according to crack width (0.3~0.5mm: single repair; >0.5mm: multiple cycle repair); 3. Energy allocation: if the energy storage capacity is less than 20%, only the repair of the key area is triggered, and a command is sent to the target module via the LoRaWAN protocol (i.e., the low power wide area network communication protocol) to trigger the photoresponsive material repair or start the UV lamp for assistance.

[0103] In addition, secondary monitoring is conducted within one hour after repair. Crack sensors continuously collect data on the closure status and calculate the repair efficiency (closure rate = repaired width / original width × 100%). Local incremental learning: if the repair efficiency is <80%, the current environmental parameters and repair results are used as new samples to update the local model weights. Cloud collaboration: data from each gateway is synchronized to the cloud monthly, the global model is optimized and distributed to edge nodes. If the same module triggers repair ≥3 times within 72 hours, it is identified as a "high-damage area" and an alarm is pushed to maintenance personnel, enabling them to view the 3D damage map through the cloud platform and carry portable UV lamps and replacement modules for on-site reinforcement.

[0104] Therefore, edge intelligent decision-making utilizes lightweight AI models deployed on revetment units to collect real-time data on structural deformation, cracks, illumination, and hydrology. After denoising and feature extraction, the data is input into an algorithm fusing CNN and LSTM to dynamically analyze damage risks and adjust repair thresholds (e.g., lowering the crack triggering threshold from 0.5mm to 0.3mm on rainy days). The decision results are then transmitted via the LoRaWAN protocol to trigger self-healing of photoresponsive materials or activate UV lamps for assistance. Simultaneously, the model parameters are optimized based on feedback from the repair results. The aim is to achieve an autonomous perception-analysis-response closed loop for the revetment system, reducing cloud dependence and communication latency, improving repair real-time performance and energy efficiency, adapting to dynamic needs in complex environments, and ensuring the long-term operation of the revetment structure and its ecological functions.

[0105] In summary, this invention utilizes a modular mortise and tenon structure design, enabling revetment units to be connected via mortise and tenon joints. This supports rapid assembly and partial replacement, adapting to complex terrain, reducing construction and maintenance costs, and improving flexibility and material utilization. Through dynamic photoresponsive self-healing materials, based on photosensitive polymers and nanomaterials, ultraviolet light triggers the rapid release of a repair agent to fill cracks, overcoming traditional temperature and salinity limitations and significantly improving repair efficiency and environmental adaptability. Furthermore, the piezoelectric power generation self-powered system utilizes water flow fluctuations and vibrations to drive piezoelectric ceramics for power generation, achieving energy self-sufficiency for the sensor network and ensuring stable system operation under extreme climates without external power supply. Finally, through edge intelligent dynamic decision-making, integrating a lightweight AI model, this invention analyzes multi-source data in real time and dynamically adjusts repair thresholds, optimizing repair timing and modes, reducing cloud dependence, and improving real-time response and energy efficiency.

[0106] The above-mentioned solution of the present invention is applicable to ecological restoration projects of urban rivers, lakes and other water bodies, and has the advantages of high efficiency, economy and environmental friendliness. Specifically, it includes the following advantages: (1) Stronger environmental adaptability: The light-response restoration mechanism breaks through the environmental temperature and salinity limitations and is applicable to a variety of water environments; (2) Zero external energy supply: Piezoelectric power generation achieves energy self-sufficiency and ensures the continuous operation of the system in extreme environments; (3) It has both ecological enhancement and water purification functions. Ecological collaborative design simultaneously strengthens the functions of bank protection, water purification and biological habitat; (4) Reduced construction and maintenance costs: The modular structure improves construction flexibility, reduces material waste, and saves maintenance costs by replacing parts locally.

[0107] Regarding the self-repair method for the ecological revetment system provided in the foregoing embodiments, this invention provides a self-repair device for the ecological revetment system, see [link / reference]. Figure 4 The diagram shows a self-repairing device for an ecological bank protection system. The device includes the following components:

[0108] The data acquisition module 402 acquires historical crack data of the ecological bank protection system, structural data, environmental data and biological detection data fed back by the low-power sensor network, as well as energy data fed back by the piezoelectric power generation mid-layer, and determines the input data with a unified data structure by aligning each data according to the timestamp.

[0109] Data analysis module 404 performs feature extraction processing on the input data, determines the model input vector, and sends the model input vector to the lightweight AI model for dynamic threshold prediction and mode selection processing to determine the repair trigger threshold and repair mode.

[0110] The repair control module 406, when detecting that the crack width collected by the crack sensor is not less than the repair trigger threshold, sends a repair command to the ultraviolet lamp through the low power wide area network communication protocol, so that the ultraviolet lamp irradiates the dynamic photoresponsive material to release the repair agent to repair the crack.

[0111] The self-repairing device for the ecological bank protection system provided in this application embodiment can significantly improve the environmental adaptability of the bank protection system.

[0112] In one embodiment, after determining the input data with a unified data structure, the data analysis module 404 is further configured to: perform noise reduction processing on the input data by using a sliding window averaging method, and perform temperature drift calibration on the noise-reduced data to determine the target input data.

[0113] In one embodiment, when performing feature extraction processing on the input data to determine the model input vector, the data analysis module 404 is further configured to: perform feature extraction processing on the crack propagation rate, illumination trend and water flow impact frequency in the target input data to determine the feature vector; use a preset Z-Score normalization processing model and a preset maximum-minimum scaling model to standardize the feature vector; and perform data fusion processing on the standardized feature vector to determine the model input vector.

[0114] In one embodiment, when sending the model input vector to the lightweight AI model for dynamic threshold prediction and mode selection processing to determine the repair trigger threshold and repair mode, the data analysis module 404 is further configured to: use the convolutional neural network and long short-term memory network in the lightweight AI model to perform dynamic threshold prediction processing on the model input vector to determine the repair trigger threshold; perform mode selection processing based on the energy storage capacity in the energy data and the water flow velocity in the environmental data; when the energy storage capacity and water flow velocity are within the preset parameter threshold range, determine the repair mode as passive repair and use natural light to initiate crack repair; when the energy storage capacity and water flow velocity are not within the preset parameter threshold range, determine the repair mode as active repair and use ultraviolet lamps to initiate crack repair.

[0115] The device provided in this embodiment of the invention has the same implementation principle and technical effect as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0116] This invention provides a server, specifically, the server includes a processor and a storage device; the storage device stores a computer program, which, when run by the processor, executes the method described in any of the above embodiments.

[0117] Figure 5This is a schematic diagram of the structure of a server provided in an embodiment of the present invention. The server 100 includes: a processor 50, a memory 51, a bus 52 and a communication interface 53. The processor 50, the communication interface 53 and the memory 51 are connected through the bus 52. The processor 50 is used to execute executable modules, such as computer programs, stored in the memory 51.

[0118] The memory 51 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 53 (which can be wired or wireless), such as the Internet, wide area network, local area network, metropolitan area network, etc.

[0119] Bus 52 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 5 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0120] The memory 51 is used to store programs. After receiving an execution instruction, the processor 50 executes the programs. The method executed by the device for defining the flow process disclosed in any of the foregoing embodiments of the present invention can be applied to the processor 50 or implemented by the processor 50.

[0121] Processor 50 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 50 or by instructions in software form. Processor 50 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this invention. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this invention can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 51. The processor 50 reads the information in memory 51 and, in conjunction with its hardware, completes the steps of the above method.

[0122] The computer program product of the readable storage medium provided in the embodiments of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the foregoing method embodiments. For specific implementation, please refer to the foregoing method embodiments, which will not be repeated here.

[0123] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0124] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An ecological bank protection system, characterized in that, The system comprises: a light-responsive repair surface layer, a piezoelectric power generation middle layer, an eco-concrete bottom layer, and an edge smart gateway; wherein... The photoresponsive repair surface layer includes: a light absorption enhancement layer, a repair agent storage layer, and an interface bonding layer. The photoresponsive repair surface layer is used to soften the dynamic photoresponsive material when exposed to natural light or ultraviolet light from an ultraviolet lamp. The piezoelectric power generation middle layer includes: an array of piezoelectric ceramic sheets, which is used to generate electrical energy when water flow impacts or the bank protection system vibrates, in order to drive the edge smart gateway and the low-power sensor network; The bottom layer of the ecological concrete includes: a porous ecological concrete layer with a porous structure, which is used to set nitrifying bacteria and aquatic plant seeds in the pores for water purification and biological habitat treatment. The edge smart gateway is used to combine historical crack data and sensor data fed back from the low-power sensor network to dynamically adjust the repair trigger threshold. When the crack width is detected to be not less than the repair trigger threshold, a repair command is sent to the ultraviolet lamp to irradiate the dynamic photoresponsive material and release the repair agent to fill the crack.

2. The ecological bank protection system according to claim 1, characterized in that, The low-power sensor network includes: structural monitoring sensors, environmental monitoring sensors, and biological monitoring sensors; wherein... The structural monitoring sensor includes a strain sensor and a crack sensor, and the structural monitoring sensor is used to feed back structural data to the edge smart gateway. The environmental monitoring sensors include: a light intensity sensor, a water level and flow rate sensor, and a water quality sensor. The environmental monitoring sensors are used to feed back environmental data to the edge smart gateway. The biomonitoring sensor includes a biofilm activity sensor, which is used to feed back biodetection data to the edge smart gateway.

3. The ecological bank protection system according to claim 1, characterized in that, The edge intelligent gateway includes: a lightweight AI model; wherein... The lightweight AI model includes a convolutional neural network and a long short-term memory network. The lightweight AI model is used to determine the repair trigger threshold and repair mode based on the historical crack data and the sensor data.

4. A self-repair method for an ecological bank protection system, characterized in that, The method is applied to the edge smart gateway of the ecological bank protection system as described in claim 1, and the method includes: The system acquires historical crack data of the ecological bank protection system, structural data, environmental data and biological detection data fed back by the low-power sensor network, as well as energy data fed back by the piezoelectric power generation mid-layer, and determines the input data with a unified data structure by aligning each data according to the timestamp. Feature extraction processing is performed on the input data to determine the model input vector, and the model input vector is sent to a lightweight AI model for dynamic threshold prediction and mode selection processing to determine the repair trigger threshold and repair mode. When the crack width detected by the crack sensor is not less than the repair trigger threshold, a repair command is sent to the ultraviolet lamp via a low-power wide-area network communication protocol, causing the ultraviolet lamp to irradiate the dynamic photoresponsive material to release the repair agent and repair the crack.

5. The self-repair method for the ecological bank protection system according to claim 4, characterized in that, Following the step of determining the input data for the unified data structure, the following is included: The input data is denoised using a sliding window averaging method, and the denoised data is then calibrated for temperature drift to determine the target input data.

6. The self-repair method for the ecological bank protection system according to claim 4, characterized in that, The step of performing feature extraction processing on the input data to determine the model input vector includes: Feature vectors are determined by extracting features from the crack propagation rate, illumination trend, and water flow impact frequency in the target input data. The feature vectors are standardized using a preset Z-Score normalization model and a preset maximum-minimum scaling model. The standardized feature vectors are then subjected to data fusion processing to determine the model input vector.

7. The self-repair method for the ecological bank protection system according to claim 4, characterized in that, The steps of sending the model input vector to a lightweight AI model for dynamic threshold prediction and mode selection, and determining the repair trigger threshold and repair mode, include: By utilizing the convolutional neural network and long short-term memory network in the lightweight AI model, dynamic threshold prediction processing is performed on the input vector of the model to determine the repair trigger threshold; Based on the energy storage capacity in the energy data and the water flow velocity in the environmental data, a mode selection process is performed. When the energy storage capacity and the water flow velocity are within the preset parameter threshold range, the repair mode is determined to be passive repair, which uses natural light to open the crack repair. When the energy storage capacity and the water flow velocity are not within the preset parameter threshold range, the repair mode is determined to be active repair, which uses ultraviolet lamps to open the crack repair.

8. A self-repairing device for an ecological bank protection system, characterized in that, The device is applied to the edge smart gateway of the ecological bank protection system as described in claim 1, and the device comprises: The data acquisition module acquires historical crack data of the ecological bank protection system, structural data, environmental data and biological detection data fed back by the low-power sensor network, as well as energy data fed back by the piezoelectric power generation mid-layer, and determines the input data with a unified data structure by aligning each data according to the timestamp. The data analysis module performs feature extraction processing on the input data, determines the model input vector, and sends the model input vector to the lightweight AI model for dynamic threshold prediction and mode selection processing to determine the repair trigger threshold and repair mode. The repair control module, when it detects that the crack width collected by the crack sensor is not less than the repair trigger threshold, sends a repair command to the ultraviolet lamp through a low-power wide area network communication protocol, so that the ultraviolet lamp irradiates the dynamic photoresponsive material to release the repair agent to repair the crack.

9. A server, characterized in that, The method includes a processor and a memory, the memory storing computer-executable instructions executable by the processor, the processor executing the computer-executable instructions to implement the method of any one of claims 4 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions that, when invoked and executed by a processor, cause the processor to perform the method described in any one of claims 4 to 7.

Images