Virtual drill system for underwater dam inspection robots and its construction method

Abstract

The present invention relates to the field of underwater robot virtual drills, and specifically discloses a virtual drill system for an underwater dam inspection robot and a construction method thereof. The system comprises a user interaction layer, a system function layer and a data resource layer, and aims to solve the problems of complex operation and high training cost of existing underwater robots. The system realizes user interaction through parameter setting, subject selection, human-machine device interaction and a visual UI. Functional integration is realized by adopting task management and evaluation, virtual device control, scene simulation rendering and scene management modules. Data management is performed using basic data, 3D model data, drill data and scene data. The system can simulate the entire operation process of the underwater robot, covering environmental simulations such as weather, water quality, and day and night changes. Operators can use the system to simulate and drill different underwater robot dam defect inspection tasks, saving manpower and material resources while improving operational proficiency and accuracy, effectively saving training costs and reducing operational risks.

Description

Technical Field

[0001] The present invention relates to the technical field of underwater robot virtual drills, and in particular to an underwater dam inspection robot virtual drill system and a construction method thereof. Background Art

[0002] As the core component of a water conservancy project, the structural safety and stability of a hydropower station dam is directly linked to the station's operational efficiency and the safety of downstream areas. As a hydropower station ages, the concrete surface of the dam, impacted by various factors such as water flow, water pressure, and structural aging, inevitably develops defects such as cracks, seepage, and spalling. This is particularly true in the underwater area in front of the dam, where long-term direct impact from the water flow causes defects to develop and develop more rapidly, posing a significant safety hazard to the dam.

[0003] To promptly detect and address these defects and ensure the safe operation of dams, manual inspection methods have traditionally been relied upon. However, manual underwater work is not only risky and difficult, but also inefficient, making it difficult to meet the demands of modern hydropower station management. With technological advances, underwater robots are beginning to be used in underwater dam inspections. With their advantages of high precision, high efficiency, and safety, underwater robots have become an important alternative to manual underwater inspections.

[0004] However, the operational process for underwater robots, especially those used for dam inspections, is complex, encompassing multiple stages including transportation, installation and deployment, deployment, underwater operations, and recovery. Each stage presents risks due to improper operation or equipment failure. Furthermore, the underwater environment is complex and variable, with factors such as water quality, current velocity, and water depth all potentially impacting the operation of underwater robots, increasing the risk and difficulty of these operations. Furthermore, underwater robots are typically expensive, and any unforeseen loss of these robots can result in significant financial losses.

[0005] Therefore, underwater robot operators must receive professional training and, through repeated practice, master the robots' working principles, operating methods, and ability to respond to emergencies. However, the virtual drill systems currently available on the market are mostly designed for underwater robots operating in the ocean, lacking systems for inland reservoirs, particularly the underwater environments of hydropower dams. Existing systems can only conduct drills within a single scenario, or require manual addition and deletion of models. They often only simulate the underwater portion of operations and fail to cover the entire operational process of underwater dam inspection robots, making it difficult to fully enhance the comprehensive skills and adaptability of drill operators.

[0006] In addition, existing virtual drill systems also have shortcomings in scene rendering effects, especially the simulation effects of underwater environments in inland reservoirs, such as water quality, lighting, biological distribution and other factors, which are not ideal and cannot provide drill personnel with a real and realistic drill environment, thus affecting the effectiveness of the drill.

[0007] In view of the above problems, the development of a virtual drill system that can cover the entire operation process of underwater dam inspection robots, simulate the real underwater environment, and support dynamic matching of multiple scenarios is of great practical significance and urgent need for improving the operational skills and adaptability of drill personnel and reducing actual operation risks. Summary of the Invention

[0008] The technical problem to be solved by the present invention is to provide a virtual rehearsal system for underwater dam inspection robots and a construction method thereof, so as to solve the limitations of existing virtual rehearsal systems for underwater dam inspection robots, and specifically to address the problems of high risk and low efficiency of traditional manual detection methods in underwater inspection of hydropower station dams, as well as the inability of existing virtual rehearsal systems to fully cover the underwater robot operation process and the unsatisfactory rendering simulation effect.

[0009] To solve the above technical problems, the present invention adopts the following technical solution: a virtual rehearsal system for underwater dam inspection robots, comprising a user interaction layer, a system function layer, and a data resource layer:

[0010] The user interaction layer includes virtual drill parameter settings, drill subject selection, human-machine device interaction, and a visual UI interface. The human-machine device interaction is used to read the robot virtual control device input and other user input, and is associated with the virtual device control module in the system function layer. The visual UI interface includes guidance information in different modes and visualization of simulation feedback information of each virtual device;

[0011] The system function layer includes the task management and assessment module, the virtual device control module, the scene simulation rendering module, and the scene management module. The system function layer defines different method interfaces to achieve data interaction and integration between different modules in this layer and with other layers.

[0012] The data resource layer is used to store the persistent management of basic system data, 3D model data, drill data, and scene data, and is also used by other layers of the system.

[0013] In a preferred solution, the virtual drill parameter setting of the user interaction layer includes setting of drill business data and system general data; and the drill subject selection is used to select different drill subjects.

[0014] In the preferred solution, the task management and assessment evaluation module of the system function layer includes exercise mode management, behavior tree and state machine construction, behavior tree and state machine integration, and assessment points and scoring. It is a task management system based on behavior tree and state machine. The system implements task process management through sequential nodes, selection nodes and parallel nodes, and triggers assessment points for scoring according to changes in robot status and behavior.

[0015] In a preferred solution, the virtual device control module of the system function layer includes underwater robot underwater motion simulation, underwater robot device function simulation, and other device function interaction simulation, wherein:

[0016] Underwater robot underwater motion simulation is used to establish an underwater motion interaction system for the underwater robot. According to the underwater robot power system and combined with underwater dynamics analysis, a simplified mechanical model of the underwater robot is established, and the underwater motion simulation of the underwater robot is realized through C# scripts;

[0017] Underwater robot equipment function simulation, used to simulate the functions of various equipment and sensors of underwater dam inspection robots;

[0018] Interactive simulation of other device functions is used to simulate other auxiliary devices.

[0019] In a preferred solution, the virtual device control module also includes an underwater robot virtual control device, which has a robot steering joystick, a forward button, a backward button, an up button, a dive button, a button to disconnect / connect with the power float, a sonar control button, a robotic arm control button, a light control button, and a camera control button, which are used for virtual control input of the underwater robot.

[0020] In a preferred solution, the underwater machinery and equipment function simulation includes:

[0021] Virtual control of side-scan sonar, single-beam sonar, and multi-beam sonar, while simulating feedback and displaying it on the control screen UI (User Interface) component;

[0022] Sensor parameter display: simulate robot sensor parameters and display them in UI components;

[0023] Virtual control of the underwater robot arm: The underwater robot arm model is grouped into components according to the degrees of freedom of the real underwater robot arm, and the action association control is performed through C# code;

[0024] Sub-machine virtual control: Establish collision trigger logic between the sub-machine compartment and the sub-machine, and use robot-sub-machine control mode switching to achieve sub-machine control and recovery.

[0025] In the preferred solution, the scene simulation rendering module of the system functional layer includes underwater rendering based on water quality characteristics, real-time underwater color correction based on color migration algorithm, and day and night and weather rendering. The day and night and weather system rendering uses Unity Shader to achieve simulation of day and night changes and different climates; underwater rendering calculates the underwater scene screen rendering effect based on water quality data of the water area, combined with the water color biological model and the underwater lighting model, and uses the color migration algorithm for real-time correction.

[0026] In a preferred solution, the system function layer scene management module includes scene matching import, scene editing, scene saving and deletion, and scene optimization:

[0027] Scene matching import, used to quickly match the most similar scene in the scene library according to the user-set rehearsal information and import it;

[0028] Scenario editing is used to manually edit and modify parts that do not meet the drill requirements based on other modules;

[0029] Scene saving and deletion, used for users to delete typical or unsatisfactory scenes;

[0030] Scenario optimization is used by users to optimize drill scenarios.

[0031] In the preferred solution, the scene matching import of the scene management module calculates the attribute weights based on the scene feature attribute values ​​input by the user through the improved entropy weight method, calculates the overall similarity using the nearest neighbor retrieval algorithm, and quickly matches and imports the scene with the highest similarity.

[0032] In a preferred solution, the system also includes a selection of a learning mode and an assessment mode. The learning mode provides a task guidance UI component, while the assessment mode hides the task prompt UI component and triggers the task assessment logic.

[0033] In the preferred solution, the system can simulate the standard operating procedures of underwater robots, including transportation, installation and deployment, hanging, separation of the surface dynamic buoy and the underwater robot, robot detection, recovery, and typical underwater defect handling procedures, such as initial inspection, precision inspection, detailed inspection, plane precision inspection, and plane detailed inspection.

[0034] A method for constructing a virtual rehearsal system for an underwater dam inspection robot based on any one of the above-mentioned methods is characterized by including setting rehearsal parameters, selecting rehearsal subjects, performing human-machine equipment interaction and visual UI display, managing and evaluating the rehearsal process through a task management and assessment evaluation module, performing equipment simulation control using a virtual equipment control module, simulating the working environment through a scene simulation rendering module, and performing scene management through a scene management module.

[0035] In a preferred solution, the scene simulation rendering calculates the underwater scene screen rendering effect by setting underwater water quality parameters, combining the water color biological model and the underwater lighting model. The specific implementation includes the following steps:

[0036] Step 1: Set scene illumination L, chlorophyll concentration CHL, organic matter concentration CDOM, and suspended particle concentration SPM;

[0037] Step 2: Calculate the total absorption coefficient and total attenuation coefficient underwater , thereby calculating the attenuation effect;

[0038] Step 3: Calculate the basic water color in Shader;

[0039] Step 4: Calculate the depth of the observation point and the distance to the observed object to obtain the observed water color at the observation point;

[0040] Step 5: Use the color migration algorithm in combination with the Computer Shader to perform real-time correction of the screen water color;

[0041] Step 6: Use rendering technology to achieve visual rendering such as dynamic caustics and volumetric light.

[0042] The underwater dam inspection robot virtual drill system and construction method provided by the present invention have the following beneficial effects:

[0043] 1. The present invention solves the problem of closing failure caused by key position deviation in the operation mode of combined electrical appliances in power plants and power systems, overcoming the limitations of the prior art of closing failure and frequent manual intervention caused by key position deviation;

[0044] 2. This invention uses behavior trees and state machines to manage, trigger, and transition tasks, and sets assessment indicators to evaluate the task completion and operational proficiency of drill personnel, thereby reducing the risk of underwater operations and improving detection efficiency.

[0045] 3. This invention analyzes historical operation data based on the user-input scene feature attribute values ​​through an improved entropy weight method, calculates attribute weights and similarities, and uses a nearest neighbor retrieval algorithm to match the best scene, enabling operators to learn and train in a risk-free environment, reducing training costs.

[0046] 4. The system of the present invention covers the entire process of underwater dam inspection robot operation, including transportation, installation and deployment, hanging, testing, recovery, etc., as well as typical defect handling processes. Drill personnel can conduct comprehensive, diverse and accurate simulation training to improve their proficiency in the underwater dam inspection robot operation process;

[0047] 5. This invention uses bio-optical modeling and real-time color migration correction technology. Based on water quality data and water optical property formulas, it establishes an underwater color model for a specific water area. It also uses Unity Shader technology to render the screen color, improving the system's simulation effect on the underwater operating environment and making the drill more realistic.

[0048] 6. The system of the present invention provides a scene matching import function, which can automatically match and import the scene with the highest similarity based on the scene feature attributes set by the user, thereby improving the diversity and accuracy of the drill;

[0049] 7. This invention uses behavior trees and state machines to achieve detailed control over task management and scoring, making task management and scoring more intelligent and accurate;

[0050] 8. This invention introduces day and night and weather system rendering technology, combines it with the bio-optical model to render underwater scenes with high precision, and uses a color migration algorithm to correct the rendering results in real time, greatly improving the authenticity of the simulation effect;

[0051] 9. The system of the present invention provides two options: learning mode and assessment mode. Users can learn or assess according to their own needs, which improves the flexibility and applicability of the system;

[0052] 10. The present invention also has an intelligent feedback mechanism, which provides real-time feedback suggestions and guidance based on user operation performance, accelerating skill acquisition and improvement; at the same time, it supports multi-user online collaborative work, simulates real team cooperation scenarios, and enhances team collaboration and communication capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The present invention will be further described below with reference to the accompanying drawings and embodiments:

[0054] Figure 1 This is an architecture diagram of a virtual drill system for underwater dam inspection robots provided by the present invention;

[0055] Figure 2 Implementing a logic diagram for the task management system of the present invention;

[0056] Figure 3 Schematic diagram of the interaction between the underwater robot virtual manipulation device and the system of the present invention;

[0057] Figure 4 This is a schematic diagram of a color transfer algorithm based on a related color space;

[0058] Figure 5 This is a workflow diagram for underwater scene rendering of the present invention;

[0059] Figure 6 Schematic diagram of the workflow of the scene management module of the present invention. DETAILED DESCRIPTION

[0060] The technical solutions of the present invention are further described below with reference to the accompanying drawings and embodiments:

[0061] Example 1

[0062] like Figure 1 and Figure 3 As shown in the figure, the underwater dam inspection robot virtual drill system includes the user interaction layer, the system function layer, and the data resource layer:

[0063] The user interaction layer includes virtual drill parameter settings, drill subject selection, human-machine device interaction, and a visual UI interface. The human-machine device interaction is used to read the robot virtual control device input and other user input, and is associated with the virtual device control module in the system function layer. The visual UI interface includes guidance information in different modes and visualization of simulation feedback information of each virtual device;

[0064] The system function layer includes the task management and assessment module, the virtual device control module, the scene simulation rendering module, and the scene management module. The system function layer defines different method interfaces to achieve data interaction and integration between different modules in this layer and with other layers.

[0065] The data resource layer is used to store the persistent management of basic system data, 3D model data, drill data, and scene data, and is also used by other layers of the system.

[0066] In this embodiment, the virtual drill parameter setting of the user interaction layer includes setting of drill business data and system general data; and the drill subject selection is used to select different drill subjects.

[0067] Furthermore, the task management and assessment evaluation module of the system functional layer includes exercise mode management, behavior tree and state machine construction, behavior tree and state machine integration, and assessment points and scoring. It is a task management system based on behavior tree and state machine. The system implements task process management through sequential nodes, selection nodes and parallel nodes, and triggers assessment points for scoring according to changes in robot status and behavior.

[0068] Furthermore, the virtual device control module of the system function layer includes underwater robot underwater motion simulation, underwater robot device function simulation, and other device function interaction simulation, wherein:

[0069] Underwater robot underwater motion simulation is used to establish an underwater motion interaction system for the underwater robot. According to the underwater robot power system and combined with underwater dynamics analysis, a simplified mechanical model of the underwater robot is established, and the underwater motion simulation of the underwater robot is realized through C# scripts;

[0070] Underwater robot equipment function simulation, used to simulate the functions of various equipment and sensors of underwater dam inspection robots;

[0071] Interactive simulation of other device functions is used to simulate other auxiliary devices.

[0072] Furthermore, the virtual device control module also includes an underwater robot virtual control device, which has a robot steering joystick, a forward button, a backward button, an up button, a dive button, a button to disconnect / connect with the power float, a sonar control button, a robotic arm control button, a light control button, and a camera control button, which are used for virtual control input of the underwater robot.

[0073] Furthermore, the underwater machine equipment function simulation includes:

[0074] Virtual control of side-scan sonar, single-beam sonar, and multi-beam sonar, while simulating feedback information and displaying it on the control screen UI component;

[0075] Sensor parameter display: simulate robot sensor parameters and display them in UI components;

[0076] Virtual control of the underwater robot arm: The underwater robot arm model is grouped into components according to the degrees of freedom of the real underwater robot arm, and the action association control is performed through C# code;

[0077] Sub-machine virtual control: Establish collision trigger logic between the sub-machine compartment and the sub-machine, and use robot-sub-machine control mode switching to achieve sub-machine control and recovery.

[0078] Furthermore, the scene simulation rendering module of the system functional layer includes underwater rendering based on water quality characteristics, real-time underwater color correction based on color migration algorithm, and day and night and weather rendering. The day and night and weather system rendering realizes the simulation of day and night changes and different climates through Unity Shader; underwater rendering calculates the underwater scene screen rendering effect based on water quality data of the water area, combined with the water color biological model and the underwater lighting model, and uses the color migration algorithm for real-time correction.

[0079] Furthermore, the system functional layer scene management module includes scene matching import, scene editing, scene saving and deletion, and scene optimization:

[0080] Scene matching import, used to quickly match the most similar scene in the scene library according to the user-set rehearsal information and import it;

[0081] Scenario editing is used to manually edit and modify parts that do not meet the drill requirements based on other modules;

[0082] Scene saving and deletion, used for users to delete typical or unsatisfactory scenes;

[0083] Scenario optimization is used by users to optimize drill scenarios.

[0084] Furthermore, the scene matching import of the scene management module calculates the attribute weights by improving the entropy weight method according to the scene feature attribute values ​​input by the user, and calculates the overall similarity by using the nearest neighbor retrieval algorithm, so as to quickly match and import the scene with the highest similarity.

[0085] Furthermore, the system also includes the selection of learning mode and assessment mode. The learning mode provides a task guidance UI component, while the assessment mode hides the task prompt UI component and triggers the task assessment logic.

[0086] Furthermore, the system can simulate the standard operating procedures of underwater robots, including transportation, installation and deployment, hanging, separation of the surface dynamic buoy and the underwater robot, robot detection, recovery, and typical underwater defect handling procedures, such as initial inspection, precision inspection, detailed inspection, plane precision inspection, and plane detailed inspection.

[0087] Example 2

[0088] In another preferred embodiment, based on the above embodiment 1, Figures 1 to 6 As shown, a method for constructing a virtual rehearsal system for an underwater dam inspection robot based on any one of the above-mentioned items is characterized in that it includes setting rehearsal parameters, selecting rehearsal subjects, performing human-machine equipment interaction and visual UI display, managing and evaluating the rehearsal process through a task management and assessment evaluation module, performing equipment simulation control using a virtual equipment control module, simulating the working environment through a scene simulation rendering module, and performing scene management through a scene management module.

[0089] In this embodiment, the scene simulation rendering calculates the underwater scene screen rendering effect by setting underwater water quality parameters, combining the water color biological model and the underwater lighting model. The specific implementation includes the following steps:

[0090] Step 1: Set scene illumination L, chlorophyll concentration CHL, organic matter concentration CDOM, and suspended particle concentration SPM;

[0091] Step 2: Calculate the total absorption coefficient and total attenuation coefficient underwater , thereby calculating the attenuation effect;

[0092] Step 3: Calculate the basic water color in Shader;

[0093] Step 4: Calculate the depth of the observation point and the distance to the observed object to obtain the observed water color at the observation point;

[0094] Step 5: Use the color migration algorithm and combine it with Computer Shader to correct the screen water color in real time;

[0095] Step 6: Use rendering technology to achieve visual rendering such as dynamic caustics and volumetric light.

[0096] Example 3

[0097] In another preferred embodiment, based on the above-mentioned embodiments 1 and 2, in order to more comprehensively describe the underwater dam inspection robot virtual rehearsal system, implementation methods, and technical features proposed in the present invention, a specific embodiment 3 is described in detail below.

[0098] 1. System Architecture

[0099] like Figure 1 As shown, the underwater dam inspection robot virtual drill system of the present invention includes a user interaction layer, a system function layer and a data resource layer.

[0100] 1) User interaction layer

[0101] The user interaction layer is the interface for interaction between the system and the user, specifically including the following aspects:

[0102] (1) Virtual drill parameter settings: Users can set the parameters required for the drill on this interface, including weather (such as sunny, rainy, foggy, etc.), water quality parameters (such as chlorophyll concentration, organic matter concentration, suspended particle concentration, etc.), water level, sound effects, and other system general data. The settings of these parameters directly affect the subsequent virtual drill environment.

[0103] (2) Drill subject selection: Users can select different drill subjects, such as standard inspection tasks, typical defect handling tasks, etc. The system will automatically load the corresponding drill scenarios and task processes based on the user's selection.

[0104] (3) Human-machine device interaction: Users interact with the virtual control device through this interface, such as reading the input of the robot's operating device (such as the steering joystick, forward button, reverse button, etc.) and other user inputs. The system associates these inputs with the virtual device control module to achieve control of the underwater robot.

[0105] (4) Visual UI: The system provides a visual user interface that displays guidance information in different modes and feedback information from each virtual device simulation. Users can intuitively understand the current drill status and robot status through the UI.

[0106] 2) System function layer

[0107] The system function layer is the core part of the system, including the task management and evaluation module, the virtual device control module, the scene simulation rendering module and the scene management module.

[0108] (1) Task management and evaluation module

[0109] like Figure 2 As shown, the task management system establishes a sequential node list based on the main task flow and sets up different selection nodes and parallel nodes to handle different task scenarios and parallel tasks. The system also establishes an underwater robot state machine to monitor the robot's state and behavior changes and design corresponding behaviors and actions based on state transitions. Furthermore, the system develops detailed task guidelines and assessment points based on the task logic and requirements, which are used to score the trainees' task completion.

[0110] (2) Virtual device control module

[0111] like Figure 3 As shown in the figure, this module uses underwater robot virtual control devices to simulate underwater robot motion and device functions. The control devices include the robot's steering joystick, forward and reverse buttons, and other components. The system controls these components through C# scripts and simulates the functions of devices such as sonar and robotic arms.

[0112] (3) Scene simulation rendering module

[0113] This module is responsible for simulating and rendering the weather and underwater environments of specific reservoirs within the operational scenario. The day / night system utilizes Unity shaders to achieve the day / night effect of the sky sphere. The weather system simulates rain, snow, and other weather effects using a particle system and post-processing techniques. Underwater scene rendering uses a bio-optical model to calculate the total attenuation coefficient and attenuation effect, and shaders are used to calculate the underwater screen color rendering results. Furthermore, the system employs a color migration algorithm and computer shaders for real-time correction of the rendering results to ensure the realism of the underwater scenes.

[0114] (4) Scene management module

[0115] like Figure 1 and Figure 6 As shown in the figure, the scenario management module includes scene matching import, scene editing, and scene saving and deletion. Based on the scene attribute values ​​entered by the user, the system uses an improved entropy weight method and nearest neighbor search algorithm to quickly match and import the most similar scenes. Users can also edit and modify scenes to meet different drill needs.

[0116] 3) Data resource layer

[0117] The data resource layer is used to persistently manage the system's basic data, 3D model data, drill records and process-related data, and drill scenario data. This data supports data interaction and integration between various layers of the system, ensuring the stability and reliability of system operation.

[0118] The system can effectively simulate the operating environment and processes of underwater dam inspection robots, improving the operational proficiency and safety of drill personnel.

[0119] Example 4

[0120] In another preferred embodiment, based on the above embodiments 1 to 3, Figure 1 As shown, the system of Example 4 includes a user interaction layer, a system function layer, and a data resource layer. The user interaction layer includes parameter setting for the underwater dam inspection robot virtual drill task, drill subject selection, human-machine device interaction, and a visual UI. The system function layer includes the task management and assessment evaluation module, virtual device control module, scene simulation rendering module, and scene management module of the underwater dam inspection robot drill system, and realizes integration and docking between modules by implementing interfaces between modules. The data resource layer includes basic data, 3D model data, drill data, and scene data, and is used to persistently manage the system's basic data and business data.

[0121] The task management and evaluation module is suitable for task control and exercise process evaluation of different operation processes of underwater dam inspection robots. The specific implementation process is as follows: Figure 2 . The task management part establishes a sequential node list according to the main task process, and establishes different selection nodes for different task scenarios, such as selecting different detection methods and different exception handling methods in different task scenarios, and setting parallel nodes for parallel tasks so as to monitor and manage parallel tasks at the same time. An underwater robot state machine is established to switch the robot's state, such as inspection state, exception handling state, etc. In the process of establishing the state machine, it is necessary to consider the transition conditions between different states, such as whether the robot has completed the task of the previous stage, whether an abnormal situation has been detected, etc., and design corresponding behaviors and actions according to the state transition. In the process of establishing the state machine, it is necessary to consider the transition conditions between different states, such as whether the robot has completed the task of the previous stage, whether an abnormal situation has been detected, etc., and design corresponding behaviors and actions according to the state transition.

[0122] This embodiment features a task management and scoring system based on behavior trees and state machines. The scoring component develops detailed task guidelines based on task logic and requirements, including specific tasks and execution requirements for each stage. Key nodes or indicators for assessment are identified as the basis for evaluating the underwater robot's performance. Task guidance nodes are added to the behavior tree to guide the underwater robot in executing tasks according to the specified task logic. Assessment states are defined in the state machine, with corresponding assessment points triggered based on the underwater robot's behavior and performance.

[0123] In addition, this embodiment also provides a choice between learning mode and assessment mode. When the user selects the learning mode, a task guidance UI component is provided to the user, and the task assessment logic is not triggered; when the user selects the assessment mode, the task prompt UI component is hidden, and the task assessment logic is triggered.

[0124] The virtual device interactive control module of this embodiment includes using the underwater robot virtual control device to simulate the underwater robot's underwater motion, simulate the underwater robot's device functions and interact with it. The interaction diagram is as follows: Figure 3 , and other device function simulation and interaction.

[0125] The underwater robot underwater motion simulation part simplifies the mechanical modeling of each propeller propeller, in which the propulsion force of each propeller is :

[0126] (1)

[0127] in is the thruster constant, is the motor speed.

[0128] The robot is subject to water resistance in water The formula is:

[0129] (2)

[0130] in is the liquid density, is the cross-sectional area of ​​the underwater robot in the direction of movement (which can be simplified to the cross-sectional area of ​​the Collider in the direction of movement), is the drag coefficient, is the speed of the underwater robot.

[0131] According to Newton's second law, the acceleration of the underwater robot can be obtained, and thus the movement speed.

[0132] Simulation of underwater robot equipment functions, such as:

[0133] Sonar simulation: Based on the actual sonar parameters, target information within the range is returned to the GUI (Graphical User Interface) and visualized in the simulated radar chart;

[0134] Robotic arm control: Group the joint nodes and components of a real five-degree-of-freedom robotic arm, and use C# scripts to achieve associated control;

[0135] Shooting function: Use Render Texture in Unity3D and use Texture2D to read Render Texture information.

[0136] The scene simulation rendering module of this embodiment is suitable for simulating the rendering of the weather in the operation scene and the underwater environment of a specific reservoir, including underwater scene rendering and weather scene rendering.

[0137] The underwater scene rendering uses a bio-optical model, imports water quality parameters of specific water areas, calculates the total attenuation coefficient and attenuation effect in the script code, and calculates the underwater screen color rendering results in Shader.

[0138] Since there are many factors that affect water color in the real world and it is impossible to take all of them into consideration, the final simulated water color may be different from the real water color. Therefore, the system adopts an algorithm based on color transfer in the relevant color space and uses Computer Shader to achieve real-time color correction, that is, the spatial distribution of the original image itself is normalized, and then transformed to the spatial distribution of the target image through transformations such as rotation, translation, and scaling. , the formula is as follows:

[0139] (3)

[0140] Indicates translation, Indicates rotation, Indicates scaling, subscript Indicates the target image, subscript Represents the original image.

[0141] Then the SVD decomposition method is used to construct the transformation matrix, and finally the formula is used for color migration. The specific steps are as follows: Figure 4 .

[0142] The overall underwater scene rendering process is as follows Figure 5 .

[0143] Day, night and weather rendering:

[0144] Day and Night System: A sky sphere that alternates between day and night is implemented by writing a shader. The sky sphere's day and night effects are achieved by predicting the scattering of light on objects of different sizes through Rayleigh scattering and Mie scattering, and calculating the atmospheric effects of the sky sphere's day and night changes.

[0145] Weather system: Use Unity3D to build particle systems such as rain, snow, and hail, and use volume components and corresponding profile files to simulate and render different weather conditions through post-processing technology.

[0146] The scene management module of this embodiment includes scene matching import, scene editing, scene saving and deletion. The specific process is as follows: Figure 6 .

[0147] Scene matching import includes analyzing historical job data based on the user-input scene feature attribute values ​​through the improved entropy weight method, calculating attribute weights, and thus calculating attribute similarity. Finally, the nearest neighbor retrieval algorithm is used to calculate the overall similarity, and the scenes are sorted according to the similarity between the user-input feature values ​​to obtain the best matching scene.

[0148] This embodiment divides the data types of the operation scene attributes into: continuous digital type, type enumeration type, ordinal type, and multi-value type. When calculating the similarity of continuous digital type attributes, the two attributes can be regarded as two points in space, and the distance between the two points is calculated using the Euclidean distance method. The closer the distance, the higher the similarity of the two attributes. Type enumeration type means that the values ​​of the attribute are of different types, and there is no mathematical relationship between each two values. In this case, the similarity can be determined by judging whether the values ​​are the same. If the values ​​are the same, the similarity is 1, otherwise it is 0. For example, the main defect types of the drill scene include "cracks, spalling, exposed reinforcement". If the main defect types of the scene are different, the similarity of the attribute is 0. Ordinal type means that the values ​​of the attribute are ordered, and the similarity is related to the difference in the order of the values. The closer the order, the higher the similarity. For example, the visibility attribute includes four orders: "clear, generally clear, very poor, and extremely poor". Multi-valued attribute means that the value of the attribute is not unique, and multiple values ​​can be selected from the option list to assign to the attribute. This embodiment uses the Jaccard distance method to calculate, dividing the size of the intersection of two sets of attributes by the size of the union.

[0149] In the preferred solution, the virtual drill parameter settings of the user interaction layer include the settings of drill business data and system general data; the drill subject selection is used to select different drill subjects; the above settings are all implemented through a graphical interface, and users can intuitively adjust parameter values ​​and preview the setting effects; at the same time, the system supports custom drill subjects, and users can flexibly add or modify them according to actual needs to ensure that the drill is close to actual combat needs.

[0150] In the preferred solution, the task management and assessment evaluation module of the system functional layer includes exercise mode management, behavior tree and state machine construction, behavior tree and state machine integration, and assessment points and scoring. It is a task management system based on behavior trees and state machines. The system implements task process management through sequential nodes, selection nodes, and parallel nodes, and triggers assessment points for scoring based on changes in robot status and behavior. The above settings can ensure the efficiency and accuracy of task execution, while realizing a precise assessment and feedback mechanism, providing solid data support for the continuous optimization and performance improvement of the robot.

[0151] In the preferred solution, the system also includes the selection of learning mode and assessment mode. The learning mode provides a task guidance UI component, while the assessment mode hides the task prompt UI component and triggers the task assessment logic. The above settings are intended to meet the learning and assessment needs of different users through flexible mode switching. In the learning mode, users can easily follow the instructions to complete tasks and improve their skills; while the assessment mode emphasizes independent problem solving and testing of learning outcomes.

[0152] In the preferred solution, the system can simulate the standard operating procedures of underwater robots, including transportation, installation and deployment, hanging, separation of the surface dynamic buoy and the underwater robot, robot detection, recovery, and typical underwater defect handling procedures, such as initial inspection, precision inspection, detailed inspection, plane precision inspection, and plane detailed inspection; the above settings ensure the efficiency and accuracy of underwater operations; the system also integrates an intelligent decision support module, which can dynamically adjust the operation plan according to real-time data, optimize resource allocation, and improve overall operation efficiency and safety.

[0153] In summary, the present invention provides a virtual rehearsal system for underwater dam inspection robots and a construction method thereof, which solves the limitations of existing virtual rehearsal systems for underwater dam inspection robots, specifically the problems of high risk and low efficiency of traditional manual detection methods in underwater inspection of hydropower station dams, and the inability of existing virtual rehearsal systems to fully cover the operation process of underwater robots and unsatisfactory rendering simulation effects. By providing a virtual rehearsal system for underwater dam inspection robots and a construction method thereof, a virtual rehearsal of the entire operation process of underwater dam inspection robots can be realized, thereby reducing the risks of actual operations and improving the operational proficiency and efficiency of the rehearsal personnel. The present invention comprehensively The task management system not only covers the entire process of underwater robot inspection, but also sets detailed task instructions and assessment points for each stage, realizing a comprehensive evaluation and improvement of the operation level of the drill personnel; the advanced scene simulation technology introduces the day and night and weather system rendering technology, combines the bio-optical model to render the underwater scene with high precision, and uses the color migration algorithm to correct the rendering results in real time, which greatly improves the authenticity of the simulation effect; the intelligent scene management, its scene management module adopts the improved entropy weight method and the nearest neighbor retrieval algorithm to realize the analysis of historical operation data and the calculation of scene similarity, providing users with more Convenient and accurate scene matching function; flexible virtual device control, its virtual device control module not only realizes high-precision simulation of the functions of various equipment of the underwater robot, but also provides a variety of control buttons and visual feedback, so that the drill personnel can more intuitively feel the various situations in actual operation; the multi-mode learning and assessment system provides two options: learning mode and assessment mode. Users can learn or assess according to their own needs, which improves the flexibility and applicability of the system; the solution has demonstrated high novelty and creativity in terms of operation process coverage, scene simulation, task management, equipment control, and learning and assessment modes; in addition, The system also integrates a real-time data analysis and feedback mechanism. Through big data analysis technology, it captures and analyzes key data during the drill in real time, provides immediate operational suggestions and improvement directions for drill personnel, accelerates the skill improvement process, and provides a solid technical guarantee for the safety and efficiency of underwater dam inspection work; at the same time, the system supports multi-user collaborative operations, and through cloud synchronization technology, it realizes real-time sharing and collaborative analysis of drill data, promotes communication between teams, and improves the overall collaborative ability to deal with complex situations; this innovative design not only enhances the practicality of the drill, but also promotes the intelligent and information-based development of underwater dam inspection technology.

Claims

1. The underwater dam inspection robot virtual drill system is characterized by: Including user interaction layer, system function layer, and data resource layer: The user interaction layer includes virtual drill parameter settings, drill subject selection, human-machine device interaction, and a visual UI interface. The human-machine device interaction is used to read the robot virtual control device input and other user input, and is associated with the virtual device control module in the system function layer. The visual UI interface includes guidance information in different modes and visualization of simulation feedback information of each virtual device; The system function layer includes the task management and assessment evaluation module, the virtual device control module, the scene simulation rendering module and the scene management module. The system function layer realizes the data interaction and integration between different modules of this layer and with other layers by defining different method interfaces; the task management and assessment evaluation module includes the exercise mode management, behavior tree and state machine construction, behavior tree and state machine integration and assessment points and scoring. It is a task management system based on behavior tree and state machine. The system realizes task process management through sequential nodes, selection nodes and parallel nodes, and triggers assessment points for scoring according to changes in robot status and behavior; the virtual device control module includes underwater robot underwater motion simulation, underwater robot equipment function simulation, and other equipment function interaction simulation. Motion simulation is used to establish an underwater motion interaction system for underwater robots. According to the underwater robot power system and combined with underwater dynamics analysis, a simplified mechanical model of the underwater robot is established, and the underwater motion simulation of the underwater robot is realized through C# scripts; underwater robot equipment function simulation is used to simulate the functions of various devices and sensors of underwater dam inspection robots; other equipment function interaction simulation is used to simulate other auxiliary equipment; the scene management module includes scene matching import, which is used to quickly match and import the most similar scenes in the scene library according to the user-set rehearsal information. Scene matching import is based on the user-input scene feature attribute values, and the attribute weights are calculated by the improved entropy weight method. The overall similarity is calculated using the nearest neighbor retrieval algorithm to quickly match and import the most similar scenes; The data resource layer is used to store the persistent management of basic system data, 3D model data, drill data, and scene data, and is also used by other layers of the system.

2. The underwater dam inspection robot virtual drill system according to claim 1 is characterized by: The virtual drill parameter setting of the user interaction layer includes the setting of drill business data and system general data; the drill subject selection is used to select different drill subjects.

3. The underwater dam inspection robot virtual drill system according to claim 2 is characterized by: The virtual device control module also includes an underwater robot virtual control device, which has a robot steering joystick, a forward button, a backward button, an ascending button, a diving button, a button for disengaging / engaging with a power float, a sonar control button, a robotic arm control button, a light control button, and a camera control button, which are used for virtual control input of the underwater robot.

4. The underwater dam inspection robot virtual training system according to claim 3 is characterized in that: The underwater machinery and equipment function simulation includes: Virtual control of side-scan sonar, single-beam sonar, and multi-beam sonar, while simulating feedback information and displaying it on the control screen UI component; Sensor parameter display: simulate robot sensor parameters and display them in UI components; Virtual control of the underwater robot arm: The underwater robot arm model is grouped into components according to the degrees of freedom of the real underwater robot arm, and the action association control is performed through C# code; Sub-machine virtual control: Establish collision trigger logic between the sub-machine compartment and the sub-machine, and use robot-sub-machine control mode switching to achieve sub-machine control and recovery.

5. The underwater dam inspection robot virtual drill system according to claim 4 is characterized in that: The scene simulation rendering module of the system functional layer includes underwater rendering based on water quality characteristics, real-time underwater color correction based on color migration algorithm, and day and night and weather rendering. The day and night and weather system rendering uses Unity Shader to achieve simulation of day and night changes and different climates; underwater rendering calculates the underwater scene screen rendering effect based on water quality data of the water area, combined with the water color biological model and the underwater lighting model, and uses the color migration algorithm for real-time correction.

6. The underwater dam inspection robot virtual training system according to claim 5 is characterized in that: The scene management module also includes scene editing, scene saving and deletion, and scene optimization: Scenario editing is used to manually edit and modify parts that do not meet the drill requirements based on other modules; Scene saving and deletion, used for users to delete typical or unsatisfactory scenes; Scenario optimization is used by users to optimize drill scenarios.

7. The underwater dam inspection robot virtual training system according to claim 6 is characterized by: The system also includes a selection of a learning mode and an assessment mode. The learning mode provides a task guidance UI component, while the assessment mode hides the task prompt UI component and triggers the task assessment logic.

8. The underwater dam inspection robot virtual training system according to claim 7 is characterized by: The system can simulate the standard operating procedures of underwater robots, including transportation, installation and deployment, hanging, separation of the surface dynamic buoy and the underwater robot, robot inspection, recovery, and typical underwater defect handling procedures, such as initial inspection, precision inspection, detailed inspection, plane precision inspection, and plane detailed inspection.

9. A method for constructing a virtual training system for an underwater dam inspection robot according to claim 8, characterized in that: It includes setting drill parameters, selecting drill subjects, conducting human-machine equipment interaction and visual UI display, managing and evaluating the drill process through the task management and assessment evaluation module, using the virtual device control module to simulate equipment control, simulating the working environment through the scene simulation rendering module, and managing the scene through the scene management module.

10. The method for constructing a virtual training system for underwater dam inspection robots according to claim 9, characterized in that: The scene simulation rendering is performed by setting underwater water quality parameters, combining the water color biological model and the underwater lighting model to calculate the underwater scene screen rendering effect. The specific implementation includes the following steps: Step 1: Set scene illumination L, chlorophyll concentration CHL, organic matter concentration CDOM, and suspended particle concentration SPM; Step 2: Calculate the total absorption coefficient and total attenuation coefficient underwater , thereby calculating the attenuation effect; Step 3: Calculate the basic water color in Shader; Step 4: Calculate the depth of the observation point and the distance to the observed object to obtain the observed water color at the observation point; Step 5: Use the color migration algorithm and combine it with Computer Shader to correct the screen water color in real time; Step 6: Use rendering technology to achieve dynamic caustics and volumetric light visual rendering.

Images