Intelligent measurement method and device for wading infrastructure

Abstract

The application belongs to the technical field of measurement, and relates to a kind of water-related infrastructure intelligent measurement method and device, the method comprises: based on historical drawing, BIM model and laser scanning point cloud data, prior modeling is carried out in multi-source heterogeneous environment, and measurement task dynamic adaptation is carried out;Based on the dynamic adaptation result of measurement task, a multi-modal adsorption propulsion robot is designed;Through the multi-modal adsorption propulsion robot, multi-spectrum fusion active sensing and defect enhancement identification are carried out in turbid water area;Based on the self-optimizing path planning algorithm of structural constraint, path planning and motion control are carried out;Design lightweight semantic communication protocol, and build interactive console in virtual reality mode;Intelligently fuse and analyze the collected multi-dimensional measurement data, and evaluate the state of water-related infrastructure.Not only the work efficiency is improved, but also the safety of the measurement process is ensured;It is helpful to realize accurate maintenance and prolong the service life of infrastructure.

Description

Technical Field

[0001] The present invention relates to the field of measurement technology, and in particular to an intelligent measurement method and device for water-related infrastructure. Background Art

[0002] The health of water-related infrastructure, such as bridge piers, tunnel linings, and underwater pipelines, is directly related to water transportation safety, water resource utilization, and ecological and environmental protection. Given the complexity and unique nature of the environments in which these facilities operate, traditional inspection methods are often inefficient and lack accuracy. Summary of the Invention

[0003] In order to solve the above technical problems, the present invention provides an intelligent measurement method for water-related infrastructure, which adopts the following technical solutions, including:

[0004] Based on historical drawings, BIM models and laser scanning point cloud data, multi-source heterogeneous environment prior modeling is carried out to dynamically adapt the measurement tasks;

[0005] Based on the dynamic adaptation results of the measurement task, a multimodal adsorption propulsion robot is designed;

[0006] The multimodal adsorption propulsion robot is used to perform multi-spectral fusion active perception and enhanced defect recognition in turbid waters;

[0007] Autonomous optimal path planning algorithm based on structural constraints for path planning and motion control;

[0008] Design a lightweight semantic communication protocol and build an interactive console using virtual reality;

[0009] The collected multi-dimensional measurement data will be intelligently integrated and analyzed to conduct status assessment of water-related infrastructure.

[0010] Preferably, the steps of performing multi-source heterogeneous environment prior modeling and dynamic adaptation of measurement tasks based on historical drawings, BIM models and laser scanning point cloud data specifically include:

[0011] Build a 3D digital twin model based on historical drawings, BIM models, and laser scanning point cloud multi-source data;

[0012] Automatically identify flat, vertical, inclined surfaces, corners, and joints, and perform environmental semantic segmentation and region recognition;

[0013] Decompose the measurement task into parameterized objectives and generate the optimal detection strategy template.

[0014] Preferably, the step of designing a multimodal adsorption propulsion robot based on the dynamic adaptation result of the measurement task specifically includes:

[0015] Based on the dynamic adaptation results of the measurement task, a modular adsorption propulsion system was designed by integrating vector thrusters, negative pressure adsorption modules, magnetic adsorption modules and micro-tracks.

[0016] The modular adsorption propulsion system is used to monitor the posture and adsorption force of the water-related infrastructure in real time;

[0017] According to the semantic information of the environment, it can switch between autonomous floating, adsorption and crawling modes.

[0018] Preferably, the step of performing multi-spectral fusion active perception and enhanced defect recognition in turbid waters by the multimodal adsorption propulsion robot specifically includes:

[0019] Through the multimodal adsorption propulsion robot, a blue-green laser and a near-infrared light source are integrated to perform multi-spectral active structured light imaging;

[0020] Adaptive optical image enhancement is performed based on the results of multi-spectral active structured light imaging;

[0021] Fuse contour sonar, surface texture optical and internal defect magnetic detection data to perform multi-sensor data fusion and defect identification.

[0022] Preferably, the autonomous optimal path planning algorithm based on structural constraints performs the steps of path planning and motion control including:

[0023] Planning the optimal detection path based on the digital twin model and real-time positioning;

[0024] Perform adaptive motion control and disturbance suppression, resist water flow disturbances, and track sliding mode control / MPC paths;

[0025] Conduct multi-robot collaborative exploration, with multiple machines working together to cover locks and tunnel structures.

[0026] Preferably, the steps of designing a lightweight semantic communication protocol and constructing an interactive console in a virtual reality manner specifically include:

[0027] Design lightweight semantic communication protocols for edge intelligence and data semantic extraction;

[0028] Dynamically adjust the coding strength to prioritize the transmission of urgent commands;

[0029] Use virtual reality to build an interactive console.

[0030] Preferably, the step of performing intelligent fusion analysis on the collected multi-dimensional measurement data and performing status assessment on the water-related infrastructure specifically includes:

[0031] Correlate the collected multi-dimensional measurement data to extract defect growth rate characteristics;

[0032] Fusing imagery, point clouds, and time series data to predict the health status of the water infrastructure;

[0033] Automatically generate repair priority recommendation reports.

[0034] In order to solve the above technical problems, the present invention further provides an intelligent measuring device for water infrastructure, which adopts the following technical solution, including:

[0035] Modeling module, used to perform a priori modeling of multi-source heterogeneous environments based on historical drawings, BIM models and laser scanning point cloud data, and to dynamically adapt measurement tasks;

[0036] The robot module is used to design a multimodal adsorption propulsion robot based on the dynamic adaptation results of the measurement task;

[0037] A perception module, configured to perform multi-spectral fusion active perception and enhanced defect recognition in turbid waters through the multi-modal adsorption propulsion robot;

[0038] The control module is used for path planning and motion control based on the autonomous optimal path planning algorithm of structural constraints;

[0039] The interactive module is used to design a lightweight semantic communication protocol and build an interactive console using virtual reality;

[0040] The evaluation module is used to intelligently integrate and analyze the collected multi-dimensional measurement data and conduct status assessment of water-related infrastructure.

[0041] In order to solve the above technical problems, the present invention also provides a computer device, which adopts the technical solution described below, including a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of the above-mentioned intelligent measurement method for water-related infrastructure when executing the computer-readable instructions.

[0042] In order to solve the above technical problems, the present invention also provides a computer-readable storage medium, which adopts the technical solution described below. The computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the steps of the above-mentioned intelligent measurement method for water-related infrastructure are implemented.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] Priori modeling can more accurately reflect the actual conditions of water-related infrastructure, providing a solid foundation for subsequent measurement and analysis. Dynamically adapting measurement tasks ensures reliable measurement data in a variety of complex environments.

[0045] The multimodal adsorption-propelled robot's design enables accurate measurements even in turbid waters. Its multi-spectral fusion of active sensing and enhanced defect recognition effectively improves detection capabilities in harsh environments, which is crucial for timely detection and resolution of potential infrastructure issues.

[0046] Through the autonomous optimal path planning algorithm with structural constraints, the robot can efficiently perform path planning and motion control, which not only improves work efficiency but also ensures the safety of the measurement process;

[0047] The lightweight semantic communication protocol and the virtual reality-based interactive console greatly enhance the convenience and intuitiveness of operation;

[0048] The intelligent fusion analysis of multi-dimensional measurement data provides comprehensive and in-depth data support for the status assessment of water-related infrastructure, helping to achieve precise maintenance and extend the service life of infrastructure. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the solutions in the present invention, a brief introduction is given below to the drawings required for use in describing the embodiments of the present invention. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0050] Figure 1 is a flow chart of an embodiment of the intelligent measurement method for water infrastructure of the present invention;

[0051] Figure 2 It is a structural diagram of an embodiment of the intelligent measuring device for water infrastructure of the present invention;

[0052] Figure 3 It is a structural diagram of an embodiment of a computer device of the present invention. DETAILED DESCRIPTION

[0053] Unless otherwise defined, all technical and scientific terms used herein have the same meanings as commonly understood by those skilled in the art to which this invention belongs. The terms used in the specification of the application are for the purpose of describing specific embodiments only and are not intended to limit the present invention. The terms "including" and "having" and any variations thereof in the specification and claims of the present invention and the accompanying drawings are intended to cover non-exclusive inclusions. The terms "first" and "second" in the specification and claims of the present invention and the accompanying drawings are used to distinguish different objects, not to describe a specific order.

[0054] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present invention. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute a separate or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0055] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0056] It should be noted that the intelligent measurement method for water-related infrastructure provided in the embodiments of the present invention is generally performed by a multi-beam echo sounder, a sonar detection system, a high-definition camera and spectral analysis device carried by an unmanned aerial vehicle, and a remote sensing satellite and ground radar system. Accordingly, the intelligent measurement device for water-related infrastructure is generally arranged in a multi-beam echo sounder, a sonar detection system, a high-definition camera and spectral analysis device carried by an unmanned aerial vehicle, and a remote sensing satellite and ground radar system.

[0057] Multibeam echo sounders, which emit fan-shaped beams to cover the underwater area, can create high-precision underwater topographic maps in real time. They are suitable for surveying large areas of rivers, lakes, and offshore waters. Sonar detection systems, which utilize the propagation characteristics of sound waves in different media, can detect the position and shape of underwater structures. They are particularly effective for inspecting underwater facilities such as bridge piers and pipelines.

[0058] Drone technology, combined with high-definition video and spectral analysis, enables rapid assessments of water quality, plankton distribution, and surface pollution without touching the water surface. Remote sensing satellites and ground-based radar systems can monitor water changes, flooding, and shoreline erosion from a macro perspective, providing a scientific basis for the long-term maintenance and planning of water-related infrastructure.

[0059] Example 1: Please refer to Figure 1 , shows a flow chart of an embodiment of the water infrastructure intelligent measurement method of the present invention. The water infrastructure intelligent measurement method comprises the following steps:

[0060] Step S1: Based on historical drawings, BIM models and laser scanning point cloud data, multi-source heterogeneous environment prior modeling is performed to dynamically adapt the measurement task.

[0061] Historical drawings refer to electronic two-dimensional engineering drawings, which record the structure, layout and design information of the building in detail and are an important basis for engineering construction and subsequent maintenance.

[0062] A BIM model, or Building Information Model, is a three-dimensional digital representation that integrates the geometric, physical, and functional characteristics of a building throughout its lifecycle. BIM models enable information sharing among all project stakeholders, improving design efficiency, reducing errors, and supporting decision-making.

[0063] Laser scanning point cloud data is a collection of numerous 3D coordinate points acquired by a laser scanner. This data accurately reflects the surface morphology and spatial relationships of objects and is widely used in fields such as topographic mapping and building facade measurement. The high precision and high resolution of laser scanning point cloud data make it a valuable data source for 3D reconstruction and refined modeling.

[0064] In this embodiment, step S1, based on historical drawings, BIM models and laser scanning point cloud data, multi-source heterogeneous environment prior modeling and dynamic adaptation of measurement tasks may specifically include the following steps:

[0065] S11, builds a three-dimensional digital twin model based on historical drawings, BIM models and laser scanning point cloud multi-source data.

[0066] The non-rigid ICP algorithm is used to fuse point clouds and BIM models, and the geometric consistency is optimized in combination with historical drawing constraints.

[0067] Point cloud registration (ICP algorithm) is expressed as: ,in represents the rotation matrix, represents the translation vector, represents the laser point cloud, Indicates the corresponding point of the BIM model.

[0068] Based on Bayesian reasoning, multi-source data fusion of historical drawings, BIM models and laser scanning point clouds is performed, which is expressed as: ,in represents the twin model, 、 Represents point cloud and BIM data respectively, Represents a priori constraints from historical drawings.

[0069] Step S11 solves the problem of multi-source data alignment by fusing non-rigid ICP with Bayesian.

[0070] S12 automatically identifies flat, vertical, inclined surfaces, corners, and joints, and performs environmental semantic segmentation and region recognition.

[0071] Based on a hybrid approach of deep learning (PointNet++) and geometric feature analysis, point cloud semantic segmentation (PointNet++ loss function) is expressed as: ,in, represents the semantic segmentation loss value, Indicates the real category (flat, vertical, inclined, corner and joint structural parts, etc.), represents the predicted probability, represents the network parameters, represents the L2 regularization coefficient.

[0072] Extract geometric features by calculating regional curvature: ， in, is the area curvature, is the eigenvalue of the covariance matrix, when Determined to be a plane, Determined as a corner ( is the threshold).

[0073] Step S12 combines geometric features with deep learning to improve the accuracy of semantic segmentation.

[0074] S13, decomposing the measurement task into parameterized objectives and generating an optimal detection strategy template.

[0075] The optimal detection strategy refers to dynamically generating the optimal measurement solution that meets engineering needs by quantitatively evaluating indicators such as measurement accuracy, efficiency, and coverage in a multi-source heterogeneous data environment. Its mathematical expression is:

[0076] ,in: is the optimal strategy, is the strategy space, is the measurement error (such as point cloud registration error), For time cost, For resource consumption, is the weight coefficient.

[0077] By combining multi-objective optimization (NSGA-II) with template matching, the task is parameterized first. Define the measurement target set ,in Indicates location, Indicates the accuracy requirement, Indicates resolution.

[0078] During specific implementation, the optimal detection strategy can also be defined, and through a multi-objective optimization algorithm, an intelligent decision-making solution can be achieved to achieve the best balance between detection coverage, energy consumption and operational risks.

[0079] The mathematical expression of the optimal detection strategy is: ,in, : optimal strategy solution, : feasible strategy space, : The proportion of undetected areas, :Completed inspection area, : total target detection area, : total energy consumption, : the instantaneous power of the i-th module, : Comprehensive risk coefficient, : probability of occurrence of the kth type of risk event, : Corresponding risk cost weight, : Optimize weight coefficient ( ).

[0080] Strategy optimization is performed through Pareto frontier solution: ,in, Indicates the actual error, Indicates the The measurement time is 2 seconds.

[0081] Perform template matching through similarity calculation: ,in, Indicates a historical policy template. Indicates task coverage.

[0082] Step S13 can decompose the task into parameterized objectives (such as "detecting a vertical crack 10m to the left of the gate"), and generate a dynamic strategy based on multi-objective optimization to achieve adaptive allocation of measurement tasks.

[0083] Step S2: Design a multimodal adsorption propulsion robot based on the dynamic adaptation results of the measurement task.

[0084] In this embodiment, step S2, designing a multimodal adsorption propulsion robot based on the dynamic adaptation result of the measurement task, may specifically include the following steps:

[0085] S21, based on the dynamic adaptation results of the measurement task, integrates vector thrusters, negative pressure adsorption modules, magnetic adsorption modules and micro tracks to design a modular adsorption propulsion system.

[0086] The vector propulsion system is designed to: ,in, : water density (1025 kg / m³), : thrust coefficient (0.8-1.2), : thruster radius (0.05m), : Impeller speed (200-800 rad / s)

[0087] The design includes a hybrid adsorption system of negative pressure adsorption and electromagnetic adsorption, wherein the negative pressure adsorption is expressed as: ,in, : Standard atmospheric pressure (101.325kPa), : Negative pressure chamber pressure (20-80 kPa), : Effective area of ​​sealing ring (0.02 m²).

[0088] Among them, electromagnetic adsorption is expressed as: ,in, : magnetic flux density (0.5-1.2 T), : Magnetization response time (50-200 ms).

[0089] The crawler traction system is expressed as: ,in, : Static friction coefficient (0.6-0.8), : critical slip velocity (0.1 m / s).

[0090] Through step S21, a vector thruster, a negative pressure adsorption module (concrete), a magnetic adsorption module (steel structure) and a micro track (bottom surface) are integrated to construct a mobile platform that can adapt to complex underwater environments.

[0091] S22, monitoring the posture and adsorption force of the water-related infrastructure in real time through the modular adsorption propulsion system.

[0092] Perform multi-source sensor fusion: ,in:

[0093] : posture state, : control input vector.

[0094] Perform adaptive Kalman filtering: ,in, : observation Jacobian matrix, : Time-varying observation noise matrix.

[0095] The magnetic adsorption force is expressed as: ,in, : magnetic adsorption force (N), : magnetic induction intensity (T), : effective area of ​​magnetic pole (m²), : Vacuum permeability ( ).

[0096] The negative pressure adsorption force is expressed as: ,in, : Negative pressure adsorption force (N), : pressure difference between inside and outside the cavity (Pa), : Effective adsorption area (m²).

[0097] Perform closed-loop control of adsorption force: ,in, : Adsorption force error, : Disturbance observer output.

[0098] Through step S22, the posture and adsorption force can be monitored in real time to prevent slipping and overturning, thereby ensuring the safety and stability of the measurement operation process.

[0099] S23, based on the semantic information of the environment, multi-modal switching among autonomous floating, adsorption and crawling modes is performed.

[0100] Using deep reinforcement learning for decision making: ,in : generalized advantage estimator, : Random policy function.

[0101] Design state transfer logic: ,in, : pressure gradient (kPa / m), : Equivalent friction coefficient.

[0102] To perform smooth switching control: ,in, : transition function, : Switching time constant (100-300 ms).

[0103] Floating mode: This mode allows the robot to move freely in the water using vector thrusters, similar to how a fish swims. In this mode, the robot can move quickly and explore a large area.

[0104] Adsorption mode: When the robot needs to stabilize at a specific location for detailed inspection, it will tightly adhere to the surface of water-related infrastructure, such as pipes and dams, through the negative pressure adsorption module or magnetic adsorption module.

[0105] Crawling mode: When it is necessary to move along the surface of a structure for continuous inspection, the robot will use micro-tracks to crawl on the surface of the structure, similar to the way reptiles move.

[0106] The robot collects and analyzes semantic information about the environment, such as water flow velocity, the material and shape of the surface structure, and so on. Based on this information, the robot determines the most appropriate movement mode. For example, on a flat and stable surface, the robot may choose crawling mode for detailed inspection; when rapid movement or positioning in a specific area is required, it may switch to floating or adsorption mode. This multimodal switching capability enables the robot to flexibly cope with complex water-related environments, improving measurement efficiency and accuracy. By adjusting its movement strategy in real time, the robot can more effectively complete various measurement tasks, providing strong support for safety and stability assessments of water-related infrastructure.

[0107] By adopting step S23, an intelligent measurement operation that is adaptive to the environment can be achieved.

[0108] Step S3: Using the multimodal adsorption propulsion robot, active multi-spectral fusion perception and enhanced defect recognition are performed in turbid waters.

[0109] Multi-spectral fusion active perception in turbid waters refers to the full-scale detection of surface and internal defects of underwater structures by actively emitting multi-band (visible light / near infrared / acoustic) detection signals in low-visibility water environments and combining multi-sensor data fusion.

[0110] Defect enhancement recognition refers to the use of digital image processing and deep learning algorithms to perform feature enhancement and intelligent classification on the acquired multimodal sensor data, thereby achieving accurate identification and quantitative evaluation of sub-millimeter defects.

[0111] In this embodiment, step S3, performing multi-spectral fusion active perception and defect enhanced identification in turbid waters by the multimodal adsorption propulsion robot, may specifically include the following steps:

[0112] S31, through the multimodal adsorption propulsion robot, integrating blue-green laser and near-infrared light source, to perform multi-spectral active structured light imaging.

[0113] The multispectral lighting system is represented as: ,in, : Emission intensity (blue-green light: 490nm, near infrared: 850nm), : Water attenuation coefficient (blue-green light: 0.2 / m, near infrared: 3.5 / m), : Surface reflectivity.

[0114] Perform structured light encoding: ,in, : spatial carrier frequency (20-50 line / mm), : Phase modulation function.

[0115] Design of turbid media imaging model: . The light intensity ultimately captured by the camera or sensor is the actual observed result during the imaging process, which includes direct light and various scattered light components. Direct light intensity refers to the portion of light that originates from the light source, passes directly through the medium, and reaches the camera without any scattering. This portion of light intensity typically retains the most scene information, but is often greatly attenuated in turbid media due to absorption and scattering. Backscattered light intensity refers to the portion of light that is scattered back to the camera by particles in the medium when it propagates in the medium. This portion of light intensity will increase image blur and background noise, reducing image contrast and clarity. Forward scattered light intensity, which means that the light is slightly scattered on the path toward the camera but can still convey some scene information.

[0116] Step S31 can overcome the scattering effect of water and obtain high-contrast surface features.

[0117] S32, performing adaptive optical image enhancement based on the multi-spectral active structured light imaging result.

[0118] The Retinex enhancement algorithm is expressed as: ,in, : Gaussian kernel function ( ).

[0119] The Retinex enhancement algorithm performs adaptive contrast enhancement on images by simulating the color constancy mechanism of the human visual system. Gaussian kernel function The logarithm of the blurred ratio realizes the dynamic adjustment of local brightness and contrast of the image. Gaussian kernel function The parameter σ is chosen between 5 and 15, which helps smooth the image while preserving key details.

[0120] Perform wavelet domain fusion: ,in, :Adaptive weight (based on local SNR). Wavelet domain fusion uses multi-spectral information, especially visible light (VIS) and near infrared (NIR) images, through adaptive weight Achieve optimal integration of image information. Weight Dynamic adjustment based on local signal-to-noise ratio (SNR) ensures that the fused image Reduce noise interference while retaining key features.

[0121] Perform a deep descattering network: ,in, : U-Net de-scattering network, : Total variation regularization coefficient (0.01-0.1). The deep de-scattering network adopts U-Net architecture to input image Perform descattering to restore details blurred by the scattering medium. Loss function Combines the L1 norm and total variation regularization term, the former ensures the descattering result and clear image similarity, the latter through The coefficient (0.01 to 0.1) controls the smoothness of the image to avoid excessive denoising and loss of details.

[0122] Step S32 can improve the image signal-to-noise ratio and defect visibility, and improve the image quality of turbid waters in real time.

[0123] S33 integrates contour sonar, surface texture optical and internal defect magnetic detection data to perform multi-sensor data fusion and defect identification.

[0124] Cross-modal registration: ,in, :Rigid body transformation matrix (6DoF). Cross-modal registration is a key step in achieving effective fusion of multi-sensor data. By solving the rigid body transformation matrix , so that the optical data and sonar data are optimally aligned in space. This step ensures that the information from different sensors can be accurately mapped in a unified coordinate system, providing the basis for subsequent feature fusion.

[0125] Perform feature-level fusion: ,in, Feature stitching operation. Feature-level fusion is the process of integrating feature information from optical, sonar, and magnetic detection sensors. This operation leverages the information captured by various sensors to form a more comprehensive and rich feature representation. This fusion approach helps improve the accuracy and robustness of defect recognition.

[0126] Set the defect classification model to: ,in, : ReLU activation function, : Network weight matrix. The softmax classifier is combined with the deep learning network structure to effectively learn and classify the fused features. The application of ReLU activation function enhances the nonlinear expression ability of the model, and the network weight matrix and It is continuously optimized through training to improve the model's ability to identify defect categories.

[0127] Step S33 can realize three-dimensional reconstruction and intelligent diagnosis of defects.

[0128] Step S4: performing path planning and motion control based on an autonomous optimal path planning algorithm with structural constraints.

[0129] The autonomous optimal path planning algorithm with structural constraints refers to an algorithm that generates the optimal path under the following constraints: geometric constraints - gate size, tunnel curvature radius, fluid constraints - velocity field distribution , Kinematic Constraints - Minimum Turning Radius wait.

[0130] The optimal detection path involves a multi-objective optimization problem: ,in: : total task time, : Energy consumption, : Coverage rate.

[0131] Sliding mode control refers to finite-time convergence control. MPC (Model Predictive Control) refers to model predictive control.

[0132] In this embodiment, step S4, performing path planning and motion control based on the autonomous optimal path planning algorithm of structural constraints may specifically include the following steps:

[0133] S41: Plan an optimal detection path based on the digital twin model and real-time positioning.

[0134] Using the improved RRT* algorithm: ,in, : Corrected geodesic distance, : Flow velocity influencing factor, : Optimal cruising speed. The improved RRT* algorithm is introduced The function comprehensively considers the impact of path length and current environmental flow rate, thereby achieving path planning that is closer to the actual situation. Geodesic distance correction term in the function It can accurately reflect the actual distance between two points, and the flow velocity factor According to the current flow rate Dynamically adjust the path cost to ensure the safety and feasibility of the path.

[0135] Perform structural constraint modeling: ,in, : the signed distance function of the i-th structure, : Safety distance threshold (usually 0.5m). By defining the structural signed distance function and safety distance threshold , constructing a set of structural constraints This step is designed to ensure that the planned path avoids critical structures and ensures the safety of the inspection process.

[0136] Perform multi-objective optimization: ,in, : state weight, : control weight, : Coverage penalty coefficient. A multi-objective optimization is performed to achieve a balance between path length, control smoothness, and area coverage. By introducing state weights and control weights , and coverage penalty coefficient , can flexibly adjust the priority of various indicators according to actual needs, so as to obtain the optimal detection path that best meets the needs of the scene.

[0137] By adopting step S31 , a global optimal detection path can be generated.

[0138] S42 performs adaptive motion control and disturbance suppression to resist water flow disturbances and track the sliding mode control / MPC path.

[0139] The adaptive sliding mode control is designed as: ,in, : Sliding surface parameters, : boundary layer thickness, : Adaptive rate. Adaptive sliding mode control is achieved by designing the sliding surface parameters and boundary layer thickness , combined with the adaptive rate Adjusting control gains , achieving a dynamic response to the error and its derivative. This control strategy can effectively suppress water flow disturbances, ensuring that the system state quickly converges to the predetermined sliding surface and moves stably along it, thereby achieving accurate tracking of the desired path.

[0140] The MPC controller is designed to: ,in, : prediction time domain (typical value 8), : Discretize the system matrix. Model predictive control achieves smooth path tracking by predicting the state of the system in the future and optimizing the control input to minimize the deviation from the reference path and the cost of the control input. Prediction time domain Choice, system matrix, 、 The discretization of and the upper and lower limit constraints of the control input together ensure the feasibility and robustness of MPC in practical applications.

[0141] The fluid disturbance observer is designed as: ,in, : nonlinear observation function, : Adaptive gain. The fluid disturbance observer is used to estimate the flow disturbance in real time by designing a nonlinear observation function and adaptive gain , achieving accurate capture of disturbance signals. The disturbance estimate output by the observer can be used for feedforward compensation, further improving the anti-disturbance capability of the control system.

[0142] By adopting step S42, accurate path tracking and anti-disturbance can be achieved.

[0143] S43, conducts multi-robot collaborative exploration, with multiple machines working together to cover the lock and tunnel structures.

[0144] The Voronoi partition is expressed as: ,in, : workspace, : The position of robot i. Voronoi divides the workspace by , for each robot Assign a unique region , ensuring that the working areas of each robot have no overlap and no omissions, thereby maximizing exploration efficiency. This partitioning method is based on the robot position , dynamically adjust the boundaries of each area to achieve uniform coverage of the space.

[0145] The task allocation is expressed as: ,in, : Comprehensive cost, : The maximum number of tasks for a single machine. Through the optimization algorithm, a series of tasks are assigned to each robot to minimize the overall cost. This cost takes into account the task execution time and energy consumption , balanced by weights α and β. At the same time, the constraints ensure that each task is performed by only one robot, and the number of tasks undertaken by a single robot does not exceed its maximum carrying capacity. This allocation mechanism not only improves the overall exploration efficiency but also ensures the rational use of robot resources, which is the key to the collaborative work of multi-robot systems.

[0146] The model predictive control optimization problem of distributed multi-robot systems can be expressed as:

[0147] ,in, = , indicating a robot In the prediction domain The control input sequence within; is a positive integer, indicating the total number of robots working together; = , indicating the The local cost function of each robot; , represents the collision penalty gain coefficient; , represents the steepness parameter of the Sigmoid function; = , indicating a robot and The Euclidean distance of , which represents the safety collision avoidance distance threshold.

[0148] Local cost function: ,in, : state error weight matrix, : Control input weight matrix. Local cost function The construction of ensures that each robot i can perform its task efficiently and accurately. The trajectory tracking term, by accumulating the weighted sum of squared state errors, forces the robot to follow the predetermined trajectory and minimize deviations. This ensures accurate and consistent exploration. The energy consumption control term, by accumulating the weighted sum of squared control inputs, optimizes the robot's energy usage and extends its operating time. This not only improves task efficiency but also enhances the sustainability of the system.

[0149] Collision avoidance terms: ,when hour, ,when hour, ,when hour, Collision avoidance term The introduction of effectively reduces the risk of collision between robots. Dynamic adjustment, when the distance approaches the safety threshold When the collision is detected, the collision avoidance weight increases sharply, ensuring that the robot can adjust its path in time to avoid collision. This flexible and responsive collision avoidance mechanism provides a safe environment for multi-robot collaborative operations in complex environments.

[0150] By adopting step S43, distributed collaborative detection can be performed.

[0151] Step S5: Design a lightweight semantic communication protocol and build an interactive console using virtual reality.

[0152] A lightweight semantic communication protocol refers to a communication protocol that extracts semantic features of data (rather than the original data itself) for transmission. It has the following characteristics: communication load is reduced by 50-70%; it supports semantic-level fault tolerance (it can still be parsed when BER ≤ 10⁻³); and it has an end-edge-cloud collaborative architecture.

[0153] Edge intelligence refers to deploying AI models at edge nodes close to data sources. Data semantics describes the characteristic representation of the intrinsic meaning of data, such as: .

[0154] In this embodiment, step S5, designing a lightweight semantic communication protocol and constructing an interactive console using a virtual reality approach, may specifically include the following steps:

[0155] S51, design a lightweight semantic communication protocol for edge intelligence and data semantic extraction.

[0156] The semantic encoder is represented as: ,in, : original data (image / point cloud, etc.), : semantic feature vector (typical dimension 128-256), : ReLU activation function. The semantic encoder uses a neural network structure to map raw data (such as images, point clouds, etc.) into low-dimensional semantic feature vectors, which helps reduce data redundancy and improve transmission efficiency. In the encoding process, the ReLU activation function Enhanced the nonlinear expression ability of the network.

[0157] Channel adaptive coding is expressed as: ,in, : Channel capacity (Shannon's theorem), : Target bit error rate. Channel adaptive coding is based on the channel capacity and target bit error rate Dynamically adjust encoding rate , ensuring reliable data transmission under limited bandwidth.

[0158] Edge inference models can be used. Edge inference models are lightweight machine learning models deployed on edge devices to parse and process semantic feature vectors in real time. , thus enabling rapid response and intelligent decision-making. The role of the edge inference model is to reduce data transmission latency while protecting user privacy because sensitive data can be processed locally without uploading to the cloud.

[0159] Step S51 enables efficient and reliable data transmission. Semantic coding reduces data dimensionality and improves transmission efficiency. Channel-adaptive coding ensures reliable data transmission in variable channel environments. The introduction of edge inference models further enhances the system's real-time and intelligence capabilities. These measures provide solid technical support for the VR interactive console, enabling users to enjoy a smoother and more intelligent interactive experience.

[0160] S52, dynamically adjust the coding strength to prioritize the transmission of emergency commands.

[0161] Prioritize instructions: By defining a priority sequence , clearly stating that emergency commands have the highest priority, control commands second, and monitoring data the lowest. This step ensures that when channel resources are limited, emergency commands are given priority transmission opportunities, thereby ensuring system security and responsiveness.

[0162] Calculate the adaptive coding rate: ,in, : basic coding strength, : Current signal-to-noise ratio. Represents the actual coding strength, which is based on the current signal-to-noise ratio and basic coding strength Dynamic adjustment. is the signal-to-noise ratio threshold, is the priority coefficient. This mechanism increases the coding strength when the signal-to-noise ratio is high, improving the reliability and efficiency of data transmission. When the signal-to-noise ratio is low, the coding strength is reduced to reduce resource usage and ensure priority transmission of urgent commands.

[0163] Preemptive transmission is then performed, allowing high-priority packets to interrupt the transmission of low-priority packets, ensuring that urgent instructions can be processed immediately. This mechanism greatly improves the system's responsiveness and flexibility, especially in emergency situations, allowing critical information to be quickly conveyed, reducing delays and risks.

[0164] S53, uses virtual reality to build an interactive console.

[0165] Rendering a 3D scene: ,in, : the intensity of the kth light source, : Bidirectional reflectance distribution function. 3D scene rendering determines the realism and detail of the virtual environment seen by users. According to the surface normal , observation direction and light source direction To calculate the reflected light intensity at each point. In this way, the complex effects of lighting and materials on the visual effects of the scene can be simulated, thus creating a realistic virtual environment.

[0166] Build a gesture interaction model: ,in, : Hand joint features, : Operation instruction category, and The gesture interaction model can accurately identify the user's operation intention based on the user's hand posture, allowing the user to interact with the virtual console in a natural and intuitive way.

[0167] Multimodal feedback is then implemented: this is achieved by combining multiple sensory inputs, including visual, auditory, and tactile. This approach enhances user immersion and engagement, providing more comprehensive and realistic feedback during operation. For example, when a user gestures to a virtual button, they not only see the visual effect of the button being pressed, but also hear the corresponding click and even feel some tactile feedback. The combined effect of these multimodal inputs enables the VR interactive console to provide a richer and more immersive user experience.

[0168] Step S6: Perform intelligent fusion analysis on the collected multi-dimensional measurement data to conduct status assessment on water-related infrastructure.

[0169] In this embodiment, step S6, performing intelligent fusion analysis on the collected multi-dimensional measurement data to assess the status of water-related infrastructure may specifically include the following steps:

[0170] S61, correlating the collected multi-dimensional measurement data to extract defect growth rate characteristics.

[0171] Build a spatiotemporal alignment model: ,in, : space-time transformation matrix (6DoF), : point cloud data at time t, : Corresponding image features at each moment. This model ensures precise spatial alignment of data collected at different time points by minimizing the error between the spatiotemporal transformation matrix and the point cloud data and image features. This not only eliminates data deviations caused by acquisition equipment movement or environmental changes but also provides an accurate spatiotemporal reference for subsequent defect analysis.

[0172] Calculate the defect growth rate: ,in, : the area of ​​defect d at time t, : Historical average area. By comparing the difference in defect area between the current moment and the previous moment, combined with the time interval and the historical average area, the system can accurately calculate the defect growth rate. This metric is crucial for timely detection and early warning of potential quality issues, preventing them from escalating.

[0173] Then perform feature fusion: ,in, : stress gradient change rate, : Curvature change. Feature fusion organically combines multiple defect features to form a single feature vector that comprehensively reflects the defect state. The defect growth rate reflects the dynamic trend of the defect, while the stress gradient change rate and curvature change characterize the static characteristics of the defect from a mechanical and geometric perspective, respectively. This fusion approach not only enriches the representation of defect information but also provides more comprehensive data support for subsequent defect classification, identification, and assessment.

[0174] Through step S61 , the defect evolution trend can be quantified and a time-varying damage model can be established.

[0175] S62, fusing the image, point cloud and time series data to predict the health status of the water-related infrastructure.

[0176] Perform multi-source data fusion: ,in, : Image CNN features (ResNet50), : Point cloud PointNet++ features, LSTM time series features. Using the Transformer model, data features from different sources—image CNN features (extracted via ResNet50), point cloud PointNet++ features, and LSTM time series features—are effectively integrated. This step captures the inherent connections and complementary information between various data types, forming a more comprehensive and detailed data representation and providing a solid foundation for subsequent analysis.

[0177] Build a degradation prediction model: ,in, : Health index (1-100%), : material degradation coefficient, : nonlinear index (1.2-1.8), : Random noise project. Degradation prediction models can simulate the degradation of facilities over time and help predict their future health. This is important for promptly identifying potential problems and developing maintenance plans.

[0178] Quantify uncertainty: ,in, : Mean prediction network, Covariance Prediction Network. Mean and covariance prediction networks are introduced to quantify the uncertainty of the prediction results. This step provides more accurate prediction confidence intervals, helping to better understand the reliability of the prediction results and provide more robust support for decision-making.

[0179] Through step S62, infrastructure life prediction can be achieved.

[0180] S63, automatically generates a maintenance priority recommendation report.

[0181] Set up a multi-objective optimization model: ,in, : weight vector of the maintenance task, : No. The cost of a maintenance task, : No. Risk reduction benefits of a maintenance task, : No. The severity of defects in each maintenance task. A multi-objective optimization algorithm is used to sort the maintenance tasks. This algorithm takes maintenance cost into consideration. , risk reduction benefits and defect severity By minimizing the weighted sum of these objective functions, the priority ranking of each maintenance task is obtained. This process ensures that maintenance resources can be efficiently and reasonably allocated to where they are most needed.

[0182] Prioritize scoring: ,in, : Structural importance (1-5 levels), : probability of failure (0-1), :Environmental impact factors, : Maintenance cost (for normalization). The maintenance tasks are further refined and scored using the priority scoring formula. This formula combines the structural importance , failure probability and environmental factors , and divided by the square root of the repair cost , to arrive at a comprehensive score This step fully considers the actual influencing factors of the maintenance task, making the scoring results closer to the actual situation.

[0183] The multi-objective optimization algorithm provides a macro-level framework for prioritizing maintenance tasks, ensuring the rationality of overall resource allocation. Within this framework, the priority scoring formula provides a more detailed and specific assessment of maintenance tasks, enabling more precise decision-making. The combination of these two ensures comprehensive maintenance decision-making while ensuring its practicality and operability.

[0184] Based on the optimization results and priority scores, a maintenance recommendation report is automatically generated. This report not only lists the priority order of each maintenance task, but also provides detailed scoring basis and explanations, providing strong decision support for maintenance personnel.

[0185] Through step S63, maintenance resource allocation can be optimized.

[0186] The embodiment is implemented, and the beneficial effects are:

[0187] Through prior modeling, the actual condition of the water-related infrastructure can be more accurately reflected, providing a solid foundation for subsequent measurement and analysis, and dynamic adaptation of the measurement task can ensure that reliable measurement data can be obtained in various complex environments;

[0188] The design of the multi-modal adsorption propulsion robot enables accurate measurement in turbid water areas, multi-spectrum fusion active perception and defect enhancement identification, effectively improving the detection capability in harsh environments, which is crucial for timely discovery and handling of potential problems of the infrastructure;

[0189] Through the autonomous optimal path planning algorithm with structural constraints, the robot can efficiently plan the path and control the motion, not only improving the operation efficiency, but also ensuring the safety of the measurement process;

[0190] The lightweight semantic communication protocol and the virtual reality-based interactive console greatly enhance the convenience and intuitiveness of the operation;

[0191] Intelligent fusion analysis of multi-dimensional measurement data provides comprehensive and in-depth data support for the state evaluation of the water-related infrastructure, which helps to realize accurate maintenance and prolong the service life of the infrastructure.

[0192] The present application can be used in a variety of general-purpose or special-purpose computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments that include any of the above systems or devices, etc. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in a distributed computing environment, in which tasks are performed by remote processing devices connected by a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media, including storage devices.

[0193] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing related hardware using computer-readable instructions. The computer-readable instructions can be stored in a computer-readable storage medium. When executed, the program can include the processes in the above-described method embodiments. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0194] It should be understood that although the steps in the flowcharts of the accompanying drawings are shown in sequence as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some of the steps in the flowcharts of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

[0195] Example 2: Further reference Figure 2 , as a response to the above Figure 1 The present invention provides an embodiment of a water infrastructure intelligent measurement device. Figure 1 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0196] like Figure 2 As shown, the intelligent measuring device 70 for water infrastructure in this embodiment includes: a modeling module 71, a robot module 72, a perception module 73, a control module 74, an interaction module 75 and an evaluation module 76.

[0197] Modeling module 71, used to perform a priori modeling of multi-source heterogeneous environments based on historical drawings, BIM models and laser scanning point cloud data, and to perform dynamic adaptation of measurement tasks;

[0198] The robot module 72 is used to design a multimodal adsorption propulsion robot based on the dynamic adaptation results of the measurement task;

[0199] The perception module 73 is used to perform multi-spectral fusion active perception and defect enhancement recognition in turbid waters through the multi-modal adsorption propulsion robot;

[0200] A control module 74 is used for performing path planning and motion control based on an autonomous optimal path planning algorithm based on structural constraints;

[0201] Interaction module 75, used to design a lightweight semantic communication protocol and build an interactive console using virtual reality;

[0202] The evaluation module 76 is used to perform intelligent fusion analysis on the collected multi-dimensional measurement data and conduct status assessment on water-related infrastructure.

[0203] The implementation of this embodiment has the following beneficial effects:

[0204] Priori modeling can more accurately reflect the actual conditions of water-related infrastructure, providing a solid foundation for subsequent measurement and analysis. Dynamically adapting measurement tasks ensures reliable measurement data in a variety of complex environments.

[0205] The multimodal adsorption-propelled robot's design enables accurate measurements even in turbid waters. Its multi-spectral fusion of active sensing and enhanced defect recognition effectively improves detection capabilities in harsh environments, which is crucial for timely detection and resolution of potential infrastructure issues.

[0206] Through the autonomous optimal path planning algorithm with structural constraints, the robot can efficiently perform path planning and motion control, which not only improves work efficiency but also ensures the safety of the measurement process;

[0207] The lightweight semantic communication protocol and the virtual reality-based interactive console greatly enhance the convenience and intuitiveness of operation;

[0208] The intelligent fusion analysis of multi-dimensional measurement data provides comprehensive and in-depth data support for the status assessment of water-related infrastructure, helping to achieve precise maintenance and extend the service life of infrastructure.

[0209] Example 3: To solve the above technical problems, the present invention also provides a computer device. Figure 3 , Figure 3 This is a basic structural block diagram of the computer device in this embodiment.

[0210] The computer device 8 includes a memory 81, a processor 82, and a network interface 83 that are interconnected through a system bus. It should be noted that the figure only shows a computer device 8 having components of a memory 81, a processor 82, and a network interface 83, but it should be understood that it is not required to implement all the components shown, and more or fewer components may be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to a microprocessor, an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a digital signal processor (DSP), an embedded device, etc.

[0211] The computer device may be a desktop computer, notebook computer, PDA, cloud server, etc. The computer device may interact with the user via a keyboard, mouse, remote control, touchpad, or voice control device.

[0212] The memory 81 includes at least one type of readable storage medium, including flash memory, a hard drive, a multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic storage, a magnetic disk, an optical disk, etc. In some embodiments, the memory 81 may be an internal storage unit of the computer device 8, such as the hard drive or internal memory of the computer device 8. In other embodiments, the memory 81 may also be an external storage device of the computer device 8, such as a plug-in hard drive, a smart media card (SMC), a secure digital (SD) card, a flash memory card, etc. Of course, the memory 81 may also include both the internal storage unit of the computer device 8 and its external storage device. In this embodiment, the memory 81 is typically used to store the operating system and various application software installed on the computer device 8, such as computer-readable instructions for the intelligent measurement method for water infrastructure. In addition, the memory 81 can also be used to temporarily store various types of data that have been output or are about to be output.

[0213] In some embodiments, the processor 82 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip. The processor 82 is generally used to control the overall operation of the computer device 8. In this embodiment, the processor 82 is used to execute computer-readable instructions stored in the memory 81 or process data, such as executing the computer-readable instructions for the intelligent measurement method for water infrastructure.

[0214] The network interface 83 may include a wireless network interface or a wired network interface. The network interface 83 is generally used to establish a communication connection between the computer device 8 and other electronic devices.

[0215] The implementation of this embodiment has the following beneficial effects:

[0216] Priori modeling can more accurately reflect the actual conditions of water-related infrastructure, providing a solid foundation for subsequent measurement and analysis. Dynamically adapting measurement tasks ensures reliable measurement data in a variety of complex environments.

[0217] The multimodal adsorption-propelled robot's design enables accurate measurements even in turbid waters. Its multi-spectral fusion of active sensing and enhanced defect recognition effectively improves detection capabilities in harsh environments, which is crucial for timely detection and resolution of potential infrastructure issues.

[0218] Through the autonomous optimal path planning algorithm with structural constraints, the robot can efficiently perform path planning and motion control, which not only improves work efficiency but also ensures the safety of the measurement process;

[0219] The lightweight semantic communication protocol and the virtual reality-based interactive console greatly enhance the convenience and intuitiveness of operation;

[0220] The intelligent fusion analysis of multi-dimensional measurement data provides comprehensive and in-depth data support for the status assessment of water-related infrastructure, helping to achieve precise maintenance and extend the service life of infrastructure.

[0221] Example 4: The present invention also provides another embodiment, namely, providing a computer-readable storage medium, which stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to perform the steps of the above-mentioned intelligent measurement method for water-related infrastructure.

[0222] The implementation of this embodiment has the following beneficial effects:

[0223] Through prior modeling, the actual condition of the water-related infrastructure can be more accurately reflected, providing a solid foundation for subsequent measurement and analysis, and dynamic adaptation of measurement tasks can ensure reliable measurement data in various complex environments.

[0224] The design of the multi-modal adsorption propulsion robot enables accurate measurement in turbid water areas, multi-spectrum fusion active perception and defect enhancement identification effectively improve the detection capability in harsh environments, which is crucial for timely discovery and handling of potential problems of the infrastructure.

[0225] Through the autonomous optimal path planning algorithm with structural constraints, the robot can efficiently plan the path and control the motion, not only improving the work efficiency, but also ensuring the safety of the measurement process.

[0226] The lightweight semantic communication protocol and virtual reality-based interactive console greatly enhance the convenience and intuitiveness of operation.

[0227] Intelligent fusion analysis of multi-dimensional measurement data provides comprehensive and in-depth data support for the state evaluation of water-related infrastructure, which helps to achieve precise maintenance and prolong the service life of the infrastructure.

[0228] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be realized by software and the necessary general hardware platform, of course, they can also be realized by hardware, but in many cases the former is a better implementation. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes a plurality of instructions for making a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) execute the various embodiment methods of the present application.

[0229] Obviously, the above-described embodiments are only some of the embodiments of the present application, not all the embodiments, and the preferred embodiments of the present application are given in the drawings, but do not limit the patent scope of the present application. The present application can be realized in many different forms, and conversely, the purpose of providing these embodiments is to make the disclosure of the present application more thorough and comprehensive. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can modify the technical solutions described in the foregoing embodiments, or make equivalent replacements to some technical features. Any equivalent structure made by using the contents of the present application specification and drawings, directly or indirectly applied to other related technical fields, is also within the patent protection scope of the present application.

Claims

1. A method for intelligent measurement of water infrastructure, characterized in that: The steps include: Based on historical drawings, BIM models and laser scanning point cloud data, multi-source heterogeneous environment prior modeling is carried out to dynamically adapt the measurement tasks; Based on the dynamic adaptation results of the measurement task, a multimodal adsorption propulsion robot is designed; The multimodal adsorption propulsion robot is used to perform multi-spectral fusion active perception and enhanced defect recognition in turbid waters; Autonomous optimal path planning algorithm based on structural constraints for path planning and motion control; Design a lightweight semantic communication protocol and build an interactive console using virtual reality; Conduct intelligent fusion analysis of the collected multi-dimensional measurement data to assess the status of water-related infrastructure; The steps of performing multi-source heterogeneous environment prior modeling and dynamic adaptation of measurement tasks based on historical drawings, BIM models, and laser scanning point cloud data specifically include: Build a 3D digital twin model based on historical drawings, BIM models, and laser scanning point cloud multi-source data; Automatically identify flat, vertical, inclined surfaces, corners, and joints, and perform environmental semantic segmentation and region recognition; Decompose the measurement task into parameterized objectives and generate the optimal detection strategy template; The autonomous optimal path planning algorithm based on structural constraints performs the following steps for path planning and motion control: Planning the optimal detection path based on the digital twin model and real-time positioning; Perform adaptive motion control and disturbance suppression, resist water flow disturbances, and track sliding mode control / MPC paths; Conduct multi-robot collaborative exploration, with multiple robots working together to cover locks and tunnel structures; The steps of performing intelligent fusion analysis on the collected multi-dimensional measurement data and evaluating the status of water-related infrastructure specifically include: Correlate the collected multi-dimensional measurement data to extract defect growth rate characteristics; Fusing imagery, point clouds, and time series data to predict the health status of the water infrastructure; Automatically generate repair priority recommendation reports.

2. The intelligent measurement method for water-related infrastructure according to claim 1, characterized in that: The steps of designing a multimodal adsorption propulsion robot based on the dynamic adaptation results of the measurement task specifically include: Based on the dynamic adaptation results of the measurement task, a modular adsorption propulsion system was designed by integrating vector thrusters, negative pressure adsorption modules, magnetic adsorption modules and micro-tracks. The modular adsorption propulsion system is used to monitor the posture and adsorption force of the water-related infrastructure in real time; According to the semantic information of the environment, it can switch between autonomous floating, adsorption and crawling modes.

3. The intelligent measurement method for water-related infrastructure according to claim 1, characterized in that: The steps of performing multi-spectral fusion active perception and defect enhanced identification in turbid waters by the multimodal adsorption propulsion robot specifically include: Through the multimodal adsorption propulsion robot, a blue-green laser and a near-infrared light source are integrated to perform multi-spectral active structured light imaging; Adaptive optical image enhancement is performed based on the results of multi-spectral active structured light imaging; Fuse contour sonar, surface texture optical and internal defect magnetic detection data to perform multi-sensor data fusion and defect identification.

4. The intelligent measurement method for water-related infrastructure according to claim 1, characterized in that: The steps of designing a lightweight semantic communication protocol and building an interactive console using a virtual reality approach specifically include: Design lightweight semantic communication protocols for edge intelligence and data semantic extraction; Dynamically adjust the coding strength to prioritize the transmission of urgent commands; Use virtual reality to build an interactive console.

5. An intelligent measuring device for water infrastructure, characterized in that: include: Modeling module, used to perform a priori modeling of multi-source heterogeneous environments based on historical drawings, BIM models and laser scanning point cloud data, and to dynamically adapt measurement tasks; The robot module is used to design a multimodal adsorption propulsion robot based on the dynamic adaptation results of the measurement task; A perception module, configured to perform multi-spectral fusion active perception and enhanced defect recognition in turbid waters through the multi-modal adsorption propulsion robot; The control module is used for path planning and motion control based on the autonomous optimal path planning algorithm of structural constraints; The interactive module is used to design a lightweight semantic communication protocol and build an interactive console using virtual reality; The evaluation module is used to intelligently integrate and analyze the collected multi-dimensional measurement data and conduct status assessments on water-related infrastructure; The modeling module is specifically used for: Build a 3D digital twin model based on historical drawings, BIM models, and laser scanning point cloud multi-source data; Automatically identify flat, vertical, inclined surfaces, corners, and joints, and perform environmental semantic segmentation and region recognition; Decompose the measurement task into parameterized objectives and generate the optimal detection strategy template; The control module is specifically used for: Planning the optimal detection path based on the digital twin model and real-time positioning; Perform adaptive motion control and disturbance suppression, resist water flow disturbances, and track sliding mode control / MPC paths; Conduct multi-robot collaborative exploration, with multiple robots working together to cover locks and tunnel structures; The evaluation module is specifically used for: Correlate the collected multi-dimensional measurement data to extract defect growth rate characteristics; Fusing imagery, point clouds, and time series data to predict the health status of the water infrastructure; Automatically generate repair priority recommendation reports.

6. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores computer-readable instructions, and when the processor executes the computer-readable instructions, the steps of the intelligent measurement method for water-related infrastructure as described in any one of claims 1 to 4 are implemented.

7. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the intelligent measurement method for water-related infrastructure according to any one of claims 1 to 4.

Images