A method and device for constructing and integrated testing of a foot-type robot test model, an electronic device and a program product

Abstract

The application provides a method and device for constructing and integrated testing of a foot-type robot test model, an electronic device and a program product. A scene model file is generated by a data acquisition terminal and a scene modeling configuration module to build a standardized test scene. A sensor modeling configuration module generates a sensor simulation model file coupled with the environment based on the file to simulate the robot sensing function. A dynamics modeling configuration module generates a qualified dynamics model file in combination with robot basic information and scene characteristics to simulate the robot motion function. An integrated test module loads the three types of models to complete collaborative simulation testing, realize collaborative verification of the sensing and motion functions, and finally update the test process configuration to realize standardized construction and integrated testing of the test model, optimize the test process relying on the standardized model, improve the standardization and adaptability of the foot-type robot simulation test, and provide reliable support for the test.

Description

Technical Field

[0001] This application relates to the field of robotics, and more specifically, to a method, apparatus, electronic device, and program product for constructing and integrating a test model for a legged robot. Background Technology

[0002] Legged robots are widely used in disaster relief, outdoor exploration and other scenarios due to their adaptability to complex terrain. Their performance testing requires high-precision simulation models to simulate real working conditions in order to reduce the cost of physical testing during the research and development phase and shorten the research and development cycle.

[0003] Existing technologies mostly build a basic simulation environment through a 3D engine and construct motion modules with dynamics software. They only simulate a few types of sensors and use fixed dynamic calculation equations. Some solutions achieve simulation by combining multiple software programs, but they only collect 3D geometric information of the scene, configure only basic functional parameters for sensors, and the dynamic models are mainly adapted to the structure of wheeled robots without being optimized for the limb structure and ground interaction characteristics of legged robots.

[0004] The existing technology has significant drawbacks: First, scene modeling only collects three-dimensional geometric data, lacking key interactive parameters such as reflectivity and friction, and the mixing of dynamic and static elements results in poor reusability of the scene model and low adaptability to real working conditions; second, the sensor coverage is limited, the parameter configuration is incomplete, and there is no dedicated physics engine for adaptation, resulting in a large deviation between simulated data and real output, making it difficult to meet the testing requirements of the 'perception' stage; third, the dynamic model is not adapted to the limb structure of legged robots, lacks a mechanical model of the foot end effector, and the dynamic equations are fixed and inflexible, making it impossible to accurately simulate complex movements and difficult to meet the testing requirements of the 'decision and execution' stage. Summary of the Invention

[0005] The purpose of this application is to provide a method, device, electronic device and program product for constructing and integrating test models for legged robots, which are used to improve the adaptability of testing and measurement accuracy.

[0006] To achieve the above objectives, the technical solutions adopted in the embodiments of this application are as follows: In a first aspect, embodiments of this application provide a method for constructing and integrating a test model for a legged robot, the method comprising: The data acquisition terminal works with the scene modeling configuration module to generate a scene model file, wherein the scene model file is used to define the virtual scene information required for the legged robot test; The sensor modeling configuration module generates a sensor simulation model file coupled with the environment based on the scene model file, wherein the sensor simulation model file is used to define the perception function information of the legged robot; The dynamics modeling configuration module combines the basic information of the legged robot with the scene characteristics to generate a verified dynamics model file, wherein the dynamics model file is used to define the motion function information of the legged robot; The integrated testing module loads the scene model file, the sensor simulation model file, and the dynamics model file to complete the integrated collaborative simulation test. Update the test process configuration.

[0007] Optionally, the step of generating a scene model file by cooperating between the data acquisition terminal and the scene modeling configuration module includes: After the data acquisition terminal is selected and calibrated, it collects multi-dimensional data, and generates standard data files through noise reduction, deduplication and format conversion. The scene modeling configuration module verifies the standard data file, calls built-in tools to complete 3D reconstruction and topology optimization, and generates a 3D model of the scene. The scene modeling configuration module identifies the scene 3D model and separates dynamic and static elements. After adding a unique identifier, it calibrates the spatial position and generates a decoupled scene element factor set. The dynamic elements are dynamic objects in the scene 3D model, and the static elements are static objects in the scene 3D model. The scene modeling configuration module generates data tags; each data tag has a mapping relationship with the dynamic element or the static element. The scene modeling configuration module integrates the scene element factor set, exports a model file in a universally compatible format, and generates a model attribute report containing the number of elements and parameter ranges.

[0008] Optionally, the step of the sensor modeling configuration module generating a sensor simulation model file coupled with the environment based on the scene model file includes: The sensor modeling and configuration module selects from a set of multi-dimensional sensing sensors and calls up sensor-specific configuration templates that match the scene model file; The sensor modeling and configuration module determines the required and optional parameters; The sensor modeling configuration module binds the required parameters, the optional parameters, and the environmental parameters in the scene model file to generate a sensor parameter set; The sensor modeling and configuration module adapts the physics engine to the sensor parameter set; The sensor modeling configuration module imports parameters from the sensor parameter set to generate a sensor simulation model; The sensor modeling configuration module establishes a coupling relationship between the sensor simulation model and the environmental parameters in the scene model file to form the sensor simulation model file.

[0009] Optionally, the integrated testing module loads the scene model file, the sensor simulation model file, and the dynamics model file to complete the integrated co-simulation testing step, including: The integrated testing module loads the scene model file, the sensor simulation model file, and the dynamics model file; The integrated testing module performs compatibility verification on the scene model file, the sensor simulation model file, and the dynamics model file to obtain a multi-dimensional model set that passes the verification. The integrated testing module configures the transmission protocol for the multidimensional model set; The integrated testing module tests the interaction link between the multidimensional model set and the device according to the transmission protocol, and constructs a data interaction rule set corresponding to the multidimensional model set. The integrated testing module initializes the scene environment parameters and the motion state of the dynamic elements according to the environment parameters in the scene model file, starts scene environment simulation and updates it in real time, and generates a real-time simulated virtual test scene. The integrated testing module collects scene environment data and robot motion data in real time in the virtual test scenario according to the sensor simulation model file, and transmits them to the dynamic model file according to the protocol and frequency of the data interaction rule set; The integrated testing module verifies that the dynamic model file retrieves a preset combination of dynamic equations, calculates motion commands based on the collected sensor data, and drives the robot simulation model to perform obstacle-crossing actions. The integration test module records all test data in real time and generates an integration test report.

[0010] Secondly, embodiments of this application provide a legged robot test model construction and integrated testing device, a data acquisition terminal and a scene modeling configuration module, used to generate a scene model file, wherein the scene model file is used to define the virtual scene information required for legged robot testing; The sensor modeling configuration module is used to generate a sensor simulation model file coupled with the environment based on the scene model file, wherein the sensor simulation model file is used to define the perception function information of the legged robot; The dynamic modeling configuration module is used to combine the basic information of the legged robot with the scene characteristics to generate a verified dynamic model file, wherein the dynamic model file is used to define the motion function information of the legged robot; The integrated testing module is used to load the scene model file, the sensor simulation model file, and the dynamics model file to complete the integrated collaborative simulation test; and to update the configuration of the test process.

[0011] Thirdly, embodiments of this application provide an electronic device, the electronic device including a memory for storing one or more programs; and a processor. When the one or more programs are executed by the processor, the method described in the first aspect and possible implementations is implemented.

[0012] Fourthly, embodiments of this application provide a program product, wherein when the program contained in the program product is executed by a processor, it implements the method described in the first aspect and possible implementations.

[0013] Compared to existing technologies, the present application provides a method, apparatus, electronic device, and program product for constructing and integrating a test model for a legged robot. First, a data acquisition terminal and a scene modeling configuration module generate a scene model file to build a standardized test scene. Then, a sensor modeling configuration module generates a sensor simulation model file coupled with the environment based on this file, simulating the robot's perception function. Next, a dynamics modeling configuration module combines the robot's basic information with scene characteristics to generate a verified dynamics model file, simulating the robot's motion function. Finally, an integrated testing module loads the three types of models to complete collaborative simulation testing, achieving collaborative verification of perception and motion functions. Finally, the test process configuration is updated to achieve standardized construction and integrated testing of the test model. By relying on standardized models to optimize the testing process, the standardization and adaptability of legged robot simulation testing are improved, providing reliable support for its testing.

[0014] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 A flowchart illustrating a method for constructing and integrating a test model for a legged robot, as provided in an embodiment of this application; Figure 2 A flowchart illustrating another method for constructing and integrating a test model for a legged robot, provided in an embodiment of this application; Figure 3 An example diagram illustrating the construction of a scene model file provided in this application embodiment; Figure 4A flowchart illustrating another method for constructing and integrating a test model for a legged robot, provided in an embodiment of this application; Figure 5 An example diagram illustrating the construction of a sensor simulation model file provided in this application embodiment; Figure 6 A flowchart illustrating another method for constructing and integrating a test model for a legged robot, provided in an embodiment of this application; Figure 7 An example diagram illustrating the construction of a dynamic model file provided in this application embodiment; Figure 8 This is a schematic diagram of the architecture of a legged robot test model construction and integrated testing device provided in an embodiment of the present invention; Figure 9 This is a schematic structural block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0018] Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0019] In the description of this application, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set" and "connection" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0020] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0021] Please see Figure 1 , Figure 1This is a flowchart illustrating a method for constructing and integrating a test model for a legged robot, as provided in an embodiment of this application. (See attached diagram.) Figure 1 The method includes: Step S1: The data acquisition terminal works with the scene modeling configuration module to generate a scene model file.

[0022] The scene model file is used to define the virtual scene information required for testing the legged robot.

[0023] Optionally, the scene model file can be a reusable model file containing standardized parameters and attribute reports, used to define the virtual scene information required for legged robot testing after multi-dimensional data acquisition, 3D reconstruction, decoupling of dynamic and static elements, and label mapping. For example, an outdoor grassland obstacle crossing scene (virtual scene) model file, containing grass (static element) and stone (dynamic element) labels, exported in FBX format, with an attached report on the number of elements and parameter ranges. S2, the sensor modeling and configuration module generates a sensor simulation model file that is coupled with the environment based on the scene model file.

[0024] Among them, the sensor simulation model file is used to define the sensory function information of the legged robot.

[0025] Optionally, the sensor simulation model file can be a standardized model file that, based on a scene model file, has undergone sensor selection, parameter binding, and physics engine adaptation, and is coupled with scene environment parameters to define the sensory function information of the legged robot. For example, a tactile sensor simulation model, bound to the grass friction coefficient, adapted to the Vortex engine, and coupled with scene brightness and temperature parameters, defines the robot's ground contact perception function.

[0026] Optionally, the sensing function information can be adapted based on sensor selection.

[0027] For example, if a tactile sensor is selected, the corresponding sensing information includes, but is not limited to, parameters such as elastic modulus and coefficient of friction. Using the example above, by binding the coefficient of friction of the grass in the scene, the pressure sensing and friction sensing functions of the legged robot's foot when it contacts the ground can be defined. This can simulate the force feedback when the foot contacts the grass and determine the softness of the ground.

[0028] If a vision sensor is selected, the corresponding perception function information includes, but is not limited to, parameters such as resolution, focal length, and exposure time, and scene brightness parameters. It defines the legged robot's ability to recognize the contours of grass terrain and rock obstacles, as well as its distance measurement function. It can perceive the size and position of obstacles and provide visual data for obstacle crossing decisions.

[0029] If the sensor is selected as a force / torque sensor, the corresponding sensing function information includes, but is not limited to: measurement range, accuracy and other parameters, binding the mechanical parameters of the foot end effector, defining the torque sensing and force feedback functions when the robot's limbs move, and sensing the joint rotation torque and the impact force of the foot hitting the ground to ensure movement stability.

[0030] If an inertial sensor is selected, the corresponding sensing function information includes, but is not limited to, parameters such as acceleration measurement range and angular velocity accuracy. It defines the robot's own posture perception and motion state perception functions, which can sense the robot's tilt angle and movement speed when crossing obstacles, and assist in adjusting the limb movement posture.

[0031] S3, the dynamics modeling configuration module, combines the basic information of the legged robot with the characteristics of the scene to generate a verified dynamics model file.

[0032] The dynamics model file is used to define the motion function information of the legged robot.

[0033] Optionally, the dynamic model file can be a standardized model file that combines the basic information of the legged robot with the characteristics of the scene, and is used to define the robot's motion function information after parameter configuration, equation adaptation, and motion verification. For example, an 80kg hexapod robot dynamic model, including the mechanical parameters of the foot actuators, adapts to the combination of equations for obstacle-crossing scenarios, passes joint motion verification, and defines the obstacle-crossing motion function of the legged robot.

[0034] Optionally, this basic information may include the basic attributes and joint data of the legged robot.

[0035] Optionally, the sensing function information includes, but is not limited to: limb movement-related function information, motion control-related function information, motion state-related function information, and complex scene adaptation-related function information.

[0036] Taking the above-mentioned 80kg hexapod robot in an outdoor grass obstacle crossing scenario as an example, the limb movement-related functional information includes robot joint parameters (joint range of motion, rotational stiffness) and foot end effector mechanical parameters (elastic modulus, contact stiffness), which define the hexapod robot's limb swinging, joint rotation, and foot landing support motion functions. This can simulate the robot's leg extension and retraction and joint turning movements when crossing obstacles, ensuring stable contact between the feet and the grass ground.

[0037] Motion control related functional information: including the combination of dynamic equations adapted to obstacle crossing scenarios (such as "Lagrange equation + joint moment equation + ground contact mechanics equation" in the document), defining the solution logic of robot motion commands, and calculating joint torque and motion trajectory based on data such as grass friction coefficient and obstacle position collected by sensors, driving the robot to complete obstacle crossing actions.

[0038] Motion status related functional information: including basic robot motion parameters (mass 80kg, center of mass position (0,0,0.5m), inertia tensor), defining the robot's posture, speed, and force feedback functions during movement, simulating the robot's tilt angle and speed adjustment during obstacle crossing, as well as impact force feedback upon landing, to ensure motion stability.

[0039] Complex scene adaptation related functional information: including scene characteristic adaptation parameters (ground contact parameters for adapting to grass friction coefficient), defining the robot's motion adaptation capability in grass scenes, simulating the robot's anti-slip movement on soft grass surfaces, limb coordination movement when crossing stones, and adapting to the complex terrain requirements of obstacle crossing scenarios.

[0040] See also Figure 1 S4, the integrated test module loads the scene model file, sensor simulation model file, and dynamics model file to complete the integrated collaborative simulation test.

[0041] S5. Update the test process configuration.

[0042] In this process, configuration updates are used to achieve a closed-loop testing process: after integration testing, deviation analysis is used to locate problems, the corresponding model is returned to adjust parameters and retested until a qualified iterative logic is achieved.

[0043] The legged robot test model construction and integrated testing method provided in this invention first generates a scene model file defining virtual scene information using a data acquisition terminal and a scene modeling configuration module, establishing a standardized test scene foundation. Then, based on this file, a sensor modeling configuration module generates a sensor simulation model file coupled with the environment and defining perception function information, accurately simulating the robot's perception function. A dynamics modeling configuration module, combining the robot's basic information and scene characteristics, generates a verified dynamics model file defining motion function information, adapting to simulate the robot's motion function. An integrated testing module loads the three types of models to complete integrated collaborative simulation testing, achieving collaborative verification of perception and motion functions. Finally, through test process configuration updates, standardized construction and integrated testing of the test model are achieved. By relying on standardized models to optimize the testing process, the standardization and adaptability of legged robot simulation testing are improved, providing reliable methodological support for its testing.

[0044] Optionally, in existing technologies, scene data often suffers from non-standardization, resulting in confusion and poor reusability of simulated objects. To address these issues, this paper proposes a possible implementation method for constructing virtual scenes, specifically... Figure 1 On this basis, Figure 2 A flowchart illustrating another method for constructing and integrating a test model for a legged robot, as provided in this application embodiment, is shown below. Figure 2 Step S1 includes: Step S100: After selecting and calibrating the data acquisition terminal, collect multi-dimensional data, and generate standard data files through noise reduction, deduplication, and format conversion.

[0045] Optionally, this multi-dimensional data acquisition can involve simultaneously acquiring five key data types: three-dimensional geometric dimensions, reflectivity, spatial brightness, surface friction, and the speed of the moving object. For example, a laser scanner can be used to acquire three-dimensional terrain data, and a friction meter can be used to acquire the friction coefficient of grass. After processing, a standard file containing multi-dimensional data can be generated.

[0046] Step S101: The scene modeling configuration module verifies the standard data file, calls the built-in tools to complete the 3D reconstruction and topology optimization, and generates a 3D scene model.

[0047] Optionally, for standard data files that fail verification, return to step S100 to collect or reprocess the data.

[0048] Step S102: The scene modeling configuration module identifies the scene's 3D model and separates dynamic and static elements. After adding exclusive identifiers, it calibrates the spatial position and generates a decoupled set of scene element factors.

[0049] Among them, dynamic elements are dynamic objects in the scene's 3D model; static elements are static objects in the scene's 3D model, thereby achieving decoupling of dynamic and static elements: separating movable dynamic object elements and fixed static object elements in the scene, adding exclusive identifiers and ensuring no spatial position offset. Optionally, the offset error is ensured to be ≤0.5mm.

[0050] For example, separate the grass (static element) from the movable rocks (dynamic element), add exclusive identifiers, calibrate the positional deviation by 0.3mm, and generate a set of scene element factors.

[0051] Step S103: The scene modeling configuration module generates data tags.

[0052] Data labels can follow a fixed format of "feature type-parameter name-parameter value". Each data label has a mapping relationship with either dynamic or static features.

[0053] Alternatively, let's continue with the example of an 80kg hexapod robot traversing an outdoor grass obstacle course (the dynamic and static elements and parameters of the scene model file): Static element label (corresponding to the grass in the scene, the static element is identified as "S"): S-Grass-Friction coefficient-0.35.

[0054] Static element label (corresponding to the grassland in the scene, supplementary dimension): S - Grassland - Reflectivity - 0.35.

[0055] Dynamic element label (corresponding to movable stones in the scene, dynamic element is identified as "D"): D-stone-size-0.4m×0.3m×0.2m.

[0056] Dynamic element label (corresponding to movable stones in the scene, supplementing motion parameters): D - Stone - Maximum movement speed - 0.1m / s.

[0057] Static element label (corresponding to scene terrain basic parameters): S-Terrain-3D scanning accuracy-≤2mm@10m.

[0058] Step S104: The scene modeling configuration module integrates the scene element factor set, exports a model file in a universally compatible format, and generates a model attribute report containing the number of elements and parameter ranges.

[0059] Optionally, scene models can be categorized and stored on storage devices according to scene type, generating unique scene model retrieval identifiers. These standardized retrieval identifiers, uniquely generated based on "scene type - test element - timestamp," are used for rapid model retrieval and management.

[0060] Optionally, the model attribute report may include, but is not limited to: basic model information, scene element information, core parameter ranges, and model validation results.

[0061] Continuing with the example of the 80kg hexapod robot traversing obstacles on outdoor grass, the basic information of its model is as follows: 1) Scene model retrieval identifier (standardized): Outdoor grass - obstacle crossing test - 20240520143045.

[0062] Note: Follow the format of "Scene Type-Test Element-Timestamp". The scene type is outdoor grass, the test element is obstacle crossing test, and the timestamp is the exact time the report was generated, which serves as the unique identifier for model retrieval and management.

[0063] 2) Model storage path: Storage device\Legged robot test scene\Outdoor scene\Obstacle crossing test\20240520\ 3) Export format: Universally compatible FBX format (version: 2020).

[0064] 4) Model file size: 86.7MB.

[0065] 5) Topology optimization status: Redundant face count 4200 (≤5000 standard threshold), optimization qualified.

[0066] The corresponding scene element information (integrated result of element factor set) is as follows: 1) Total number of elements: 2 categories (1 category of static elements and 1 category of dynamic elements).

[0067] 2) Static element details (identifier prefix: S); Element name: Outdoor grassland terrain; Quantity of elements: 1 group (covering a test area of ​​10m×8m); Associated data tags: S-Grassland-Friction coefficient-0.35, S-Grassland-Reflectivity-0.35, S-Grassland-Brightness-5000lux, S-Terrain-3D scanning accuracy-≤2mm@10m.

[0068] 3. Dynamic Element Details (Identifier Prefix: D); Element Name: Movable Stone Obstacle; Element Quantity: 8 (evenly distributed in the test area); Associated Data Tags: D-Stone-Size-0.4m×0.3m×0.2m, D-Stone-Maximum Moving Speed-0.1m / s, D-Stone-Reflectivity-0.28.

[0069] The corresponding core parameter ranges (determined based on multi-dimensional collected data, after noise reduction and calibration) are shown in Table 1 below: Table 1

[0070] The corresponding model validation results are as follows: 1) Data integrity: Validation passed (no missing or abnormal data in multiple dimensions).

[0071] 2) Element decoupling status: Verification passed (dynamic and static elements are clearly separated, the markings are standardized, and the position calibration deviation is within the standard threshold of 0.3mm ≤ 0.5mm).

[0072] 3) Format compatibility: Verification passed (FBX format can be adapted to the subsequent sensor modeling and integration test module loading requirements).

[0073] 4) Overall conclusion: The scenario model file meets the requirements for outdoor grass obstacle crossing test of legged robots and can be used for subsequent sensor simulation model binding and integrated collaborative simulation test.

[0074] This example details the entire process of building scene model files. Multi-dimensional data acquisition overcomes the limitations of existing technologies that only collect 3D geometric data, making scene parameters more comprehensive and improving scene adaptability. Decoupling of dynamic and static elements and label mapping avoid element confusion, improving model reusability. Simultaneously, labels provide accurate association criteria for subsequent sensor parameter binding, enhancing model collaboration efficiency. Topology optimization and format standardization ensure model compatibility, adapting to different simulation platforms. Attribute reports facilitate model management and parameter traceability, further improving the standardization and operability of scene modeling.

[0075] under Figure 2 The illustrated embodiments are provided as examples. Specifically, Figure 3This is an example diagram illustrating the construction of a scene model file, provided as an embodiment of this application. See also... Figure 3 The component manufacturing process is as follows: Step S10: Construct indoor and outdoor scenes.

[0076] Indoor scenes may include, but are not limited to: living room, bedroom, stairs, etc.; outdoor scenes may include, but are not limited to: grass, sidewalk, etc.

[0077] Step S11: Set the constituent elements and parameter elements of the corresponding scene.

[0078] The components include, for example, obstacles and ground conditions; the parameters include, for example, dimensions, material parameters, and environmental parameters.

[0079] Step S12a: Collect the three-dimensional information of scene elements using a calibrated three-dimensional scanning device to form three-dimensional dimensional scanning data.

[0080] Step S12b: Collect the reflectance of scene elements using a calibrated reflectance measurement device.

[0081] Step S12c: Use a calibrated brightness measurement device to collect the spatial brightness of the scene environment.

[0082] Step S12d: Collect the speed of moving objects in the scene environment using the calibrated data.

[0083] Step S12e: Use a calibrated dynamic and static friction force measurement device to collect the friction force of the travel surface in the scene environment.

[0084] Step S13a: Import 3D model building software (e.g., CAD) to perform 3D reconstruction of scene elements and form a 3D model of the scene elements.

[0085] Step S13b: Determine the detailed distribution of reflectivity.

[0086] Step S13c: Determine the detailed distribution of spatial brightness.

[0087] Step S13d: Determine the detailed distribution of velocity.

[0088] Step S13e: Determine the detailed distribution of frictional forces on the traveling surface.

[0089] Step S14a: Extract the key structures from the 3D model and simplify them to obtain simple structural line segments.

[0090] Step S14b: Extract static and dynamic elements.

[0091] Step S15a: Decouple the static and dynamic elements in the scene.

[0092] Step S15b: Generate data tags based on the detailed distribution of reflectivity, the detailed distribution of spatial brightness, the detailed distribution of velocity, and the detailed distribution of frictional force on the traveling surface.

[0093] Step S16: Associate and map dynamic elements, static elements and data labels to form a scene element factor set.

[0094] Optionally, to address the problems of existing sensor simulations being disconnected from the scene, non-standard parameter configurations, and large simulation deviations, this application provides a possible implementation method for a sensor modeling configuration module to generate a sensor simulation model file coupled with the environment based on a scene model file. Specifically, in Figure 1 On this basis, Figure 4 A flowchart illustrating another method for constructing and integrating a test model for a legged robot, as provided in this application embodiment, is shown below. Figure 4 Step S2 includes: Step S200: The sensor modeling and configuration module selects from the multi-dimensional sensing sensors and calls the sensor-specific configuration template that matches the scene model file.

[0095] Step S201: The sensor modeling and configuration module determines the required and optional parameters.

[0096] Optionally, taking a vision sensor as an example, its mandatory parameters are resolution, focal length, etc., and optional parameters are exposure time, field of view, etc.; for a tactile sensor, its mandatory parameters are elastic modulus, coefficient of friction, etc., and optional parameters are temperature compensation range, measurement accuracy, etc.

[0097] Step S202: The sensor modeling configuration module binds the required parameters, optional parameters and environmental parameters in the scene model file to generate a sensor parameter set.

[0098] Optionally, the environmental parameters in the scene model file are retrieved to verify the rationality of the associated parameter values. If the parameters are invalid, they are modified in real time and re-verified. Once the verification is successful, a qualified set of sensor parameters is generated.

[0099] In addition, mandatory and optional parameters are bound to environmental parameters in the scene model file. This aims to ensure that sensor parameters are adapted to the scene environment, reduce simulation deviations, and improve the realism of sensor simulations. For example, the focal length parameter of the visual sensor is bound to the scene brightness parameter, and the friction coefficient of the tactile sensor is bound to the friction coefficient of the grass, generating a sensor parameter set.

[0100] Step S203: The sensor modeling and configuration module adapts the sensor parameter set to the physics engine.

[0101] For example, visual sensors are adapted to the PhysX engine, while haptic, force / torque, and inertial sensors are adapted to the Vortex engine.

[0102] Step S204: The sensor modeling configuration module imports parameters from the sensor parameter set to generate a sensor simulation model.

[0103] Optionally, preliminary functional testing can be conducted; if the test fails, the engine adaptation parameters can be adjusted. Upon successful testing, a preliminarily verified sensor simulation model is generated. Sensor-specific physics engine adaptation: a dedicated physics engine is matched according to the sensor type, optimizing the simulation accuracy of core functions.

[0104] Step S205: The sensor modeling configuration module establishes a coupling relationship between the sensor simulation model and the environmental parameters in the scene model file to form a sensor simulation model file.

[0105] Optionally, based on this example, deep coupling between the sensor and the scene environment can be achieved: the sensor model is bound to parameters such as scene temperature, humidity, brightness, and friction to simulate the output characteristics under real-world conditions.

[0106] Optionally, the sensor modeling configuration module stores model files according to sensor type and applicable scenario, generates a unique sensor model retrieval identifier, and obtains a sensor simulation model file coupled with the environment.

[0107] Continuing with the example of an 80kg hexapod robot traversing an outdoor grass obstacle course, the sensor modeling and configuration module allows for centralized selection of multi-dimensional sensing sensors, including vision, touch, force / torque, and inertia sensors, and calls upon a custom configuration template matched to the grass obstacle course scenario. It determines the parameters of various sensors: resolution and focal length are mandatory for vision sensors, with optional exposure time; elastic modulus and friction coefficient are mandatory for touch sensors, with optional temperature compensation range. These parameters are then bound to environmental parameters in the scene model to generate a standardized sensor parameter set. A physics engine is adapted to the parameter set: the PhysX engine for vision sensors and the Vortex engine for touch and other sensors. The parameters are imported to generate a preliminary sensor simulation model, which is then deeply coupled with environmental parameters such as scene temperature and brightness to form a sensor simulation model file coupled with the environment. Parameter binding follows a one-to-one rule, engine adaptation optimizes sensor simulation accuracy, and the coupling relationship ensures that the sensors can respond to scene changes in real time.

[0108] This example standardizes the construction process of sensor simulation models. Multi-dimensional sensor selection covers the perception needs of legged robots. Parameter binding and environment coupling solve the problem of sensor simulation being disconnected from the scene, reducing the deviation between simulated data and real sensor data, and improving simulation accuracy. Physics engine adaptation further optimizes the simulation effect of perception functions, ensuring that sensors can accurately reflect changes in the scene environment, providing reliable perception data for the dynamics model, and guaranteeing the smoothness of the "perception and decision-making" link. At the same time, the standardized parameter configuration process improves the efficiency and standardization of sensor modeling, adapting to the modeling needs of different scenes and different types of sensors.

[0109] The following uses visual sensors, tactile sensors, force / torque sensors, and inertial sensors as examples to illustrate... Figure 4 The illustrated embodiments are provided as examples. Specifically, Figure 5 This is an example diagram illustrating the construction of a sensor simulation model file, provided as an embodiment of this application. See also... Figure 5 The process includes: Step S20a: Determine the parameters of the vision sensor.

[0110] The parameters of this visual sensor include, but are not limited to: resolution, focal length, depth of field, and frame rate. Required parameters include resolution and focal length. Optional parameters include depth of field and frame rate.

[0111] Step S20b: Determine the parameters of the tactile sensor.

[0112] The parameters of this tactile sensor include, but are not limited to: resolution, sensitivity, elastic modulus, and coefficient of friction. Elastic modulus and coefficient of friction are mandatory parameters. Resolution and sensitivity are optional parameters.

[0113] Step S20c: Determine the parameters of the force / torque sensor.

[0114] The force / torque sensor parameters include, but are not limited to: dimensional parameters, sensitivity parameters, cross-coupling, and damping ratio. Required parameters include dimensional parameters and sensitivity parameters. Optional parameters include cross-coupling and damping ratio.

[0115] Step S20d: Determine the parameters of the inertial sensor.

[0116] The parameters of this inertial sensor include, but are not limited to: sensitivity, sampling frequency, response time, and bandwidth. Sensitivity and sampling frequency are mandatory parameters. Response time and bandwidth are optional parameters.

[0117] Step S21a: Bind environmental parameters to the vision sensor.

[0118] This environmental parameter includes, but is not limited to, contrast parameters and installation method.

[0119] Step S22b: Bind environmental parameters to the tactile sensor.

[0120] These environmental parameters include, but are not limited to: light intensity, temperature and humidity, installation method, and electromagnetic interference.

[0121] Step S22c: Bind environmental parameters to the force / torque sensor.

[0122] These environmental parameters include, but are not limited to: temperature and humidity parameters, installation method, and electromagnetic interference.

[0123] Step S22d: Bind environmental parameters to the inertial sensor.

[0124] These environmental parameters include, but are not limited to: temperature and humidity parameters, installation method, and electromagnetic interference.

[0125] Step S23a: Adapt the physics engine 1 to the visual sensor.

[0126] Step S23b: Adapt the physics engine 2 to the tactile sensor.

[0127] Step S23c: Adapt the physics engine 3 to the force / torque sensor.

[0128] Step S23d: Adapt the physics engine for the inertial sensor 4.

[0129] Step S24a: Generate a visual sensor simulation model for the visual sensor.

[0130] Step S24b: Generate a tactile sensor simulation model for the tactile sensor.

[0131] Step S24c: Generate a force / torque sensor simulation model for the force / torque sensor.

[0132] Step S24d: Generate an inertial sensor simulation model for the inertial sensor.

[0133] Step S25: Establish a coupling relationship between the sensor simulation model and the environmental parameters in the scene model file to form a sensor simulation model file.

[0134] Optionally, to address the issues of non-standardized testing processes, poor model collaboration, and abnormal data transmission in existing systems, the following provides a possible implementation method for integrating the testing module to load scene model files, sensor simulation model files, and dynamic model files to complete integrated collaborative simulation testing. Specifically, in Figure 1 On this basis, Figure 6A flowchart illustrating another method for constructing and integrating a test model for a legged robot, as provided in this application embodiment, is shown below. Figure 6 Step S4 includes: Step S400: The integrated test module loads the scene model file, sensor simulation model file, and dynamics model file.

[0135] Step S401: The integrated testing module performs compatibility verification on the scene model file, sensor simulation model file, and dynamic model file to obtain a multi-dimensional model set that has passed the verification.

[0136] The compatibility verification refers to verifying the compatibility of file formats, parameter ranges, and data interfaces. It verifies the file format compatibility and parameter range rationality of the three types of models, eliminates interface mismatch issues, and obtains a multi-dimensional model set that passes verification. This resolves problems of poor model collaboration and interface incompatibility, laying the foundation for integrated simulation testing.

[0137] Step S402: The integration test module configures the transmission protocol for the multidimensional model set.

[0138] The transmission protocol configuration can include: configuring the core TCP / IP data transmission protocol and the auxiliary UDP protocol for the three model sets, and setting the data transmission format. For example, "Sensor → Dynamics Model" can be in JSON format, and "Dynamics → Sensor Model" can be in XML format. The data transmission frequency can be set to 100Hz according to the real-time requirements of the test (for example, it can be adjusted to 200Hz under extreme conditions).

[0139] Step S403: The integration test module tests the interaction link between the multidimensional model set and the device according to the transmission protocol, and constructs the data interaction rule set corresponding to the multidimensional model set.

[0140] Optionally, the core function of this data interaction rule set is to standardize the data interaction behavior among the three types of models: tagged scene model files, sensor simulation model files coupled with the environment, and verified dynamic model files, providing a unified guideline for the collaborative simulation of the three models.

[0141] Specifically, its functions include: clarifying the protocol used for data transmission (TCP / IP core protocol and UDP auxiliary protocol), data transmission format (JSON / XML format between sensors and dynamic models), and transmission frequency (100Hz~200Hz), ensuring the real-time performance, integrity, and accuracy of data transmission between models, avoiding problems such as data packet loss, excessive latency, or format incompatibility, and ensuring the smooth implementation of integrated collaborative simulation testing of the entire "perception, decision-making, and execution" chain of legged robots.

[0142] Step S404: The integrated testing module initializes the scene environment parameters and the motion state of dynamic elements according to the environment parameters in the scene model file, starts the scene environment simulation and updates it in real time, and generates a real-time simulated virtual test scene.

[0143] For example, environmental parameters: temperature 25℃, brightness 5000 lux; dynamic element motion state: stone speed 0.5m / s.

[0144] Step S405: The integrated test module and sensor simulation model file collect scene environment data and robot motion data in real time in the virtual test scenario, and transmit them to the dynamic model file according to the protocol and frequency of the data interaction rule set.

[0145] For example, environmental data for this scenario: brightness, temperature, etc.; robot motion data: joint angles, foot contact force, etc.

[0146] Optionally, the aforementioned scene environment data and robot motion data can also be synchronously stored in the data recording system.

[0147] Step S406: The integrated test module verifies the dynamic model file by retrieving the preset combination of dynamic equations, calculates motion commands based on the collected sensor data, and drives the robot simulation model to perform obstacle-crossing actions.

[0148] Step S407: The integration test module records all test data in real time and generates an integration test report.

[0149] Optionally, after the simulation is completed, the data is categorized and converted into visualizations such as time-joint angle curves and foot contact force bar charts. These are then compared and analyzed with physical test benchmark data to generate an integrated test report.

[0150] Continuing with the example of an 80kg hexapod robot navigating an outdoor grass obstacle course, the integrated test module loads three types of model files: scene, sensor, and dynamics. Compatibility checks are performed on the file formats, parameter ranges, and data interfaces. Any discrepancies are adjusted and corrected to obtain a validated multi-dimensional model set. The model set is configured with the TCP / IP core transmission protocol and the UDP auxiliary protocol, and an interactive link test is conducted according to this protocol (continuous transmission of 1000 sets of data, no packet loss, latency ≤10ms). A data interaction rule set containing the transmission protocol, 100Hz frequency, and JSON / XML format is constructed. The scene and dynamic element states are initialized based on the scene environment parameters, and a real-time simulated virtual test scene is initiated. Scene and robot motion data are collected using the sensor simulation model and transmitted to the dynamics model according to the rule set. The dynamics model is verified to retrieve the combination of "Lagrange equations + joint moment equations + ground contact mechanics equations" to solve commands that drive the robot to overcome obstacles. All data is recorded in real time, generating an integrated test report containing test information, data statistics, and conclusions. In this process, compatibility checks eliminate interface incompatibility issues, the data interaction rule set standardizes data transmission, and the equation set is appropriately matched to the obstacle course scenario.

[0151] This example standardizes the integrated collaborative testing process. Compatibility verification ensures smooth collaboration among the three models, resolving issues of model interface incompatibility and poor collaboration. Interaction link testing and data interaction rule sets ensure the real-time, complete, and accurate transmission of data; packet loss-free and low-latency transmission guarantees smooth testing. Real-time scenario simulation and appropriate matching of multiple equation sets accurately reproduce the entire "perception-decision-execution" link in the robot's obstacle-crossing process. Test data is complete and traceable, providing a reliable basis for robot performance deviation analysis. The integrated test report standardizes the presentation of test results, facilitating subsequent process optimization, further improving the standardization and practicality of simulation testing, and reducing the cost of physical testing.

[0152] The following is about Figure 6 The illustrated embodiments are provided as examples. Specifically, Figure 7 This is an example diagram illustrating the construction of a dynamic model file, provided as an embodiment of this application. See also... Figure 7 The process includes: Step S40a: Determine the basic attributes of the legged robot.

[0153] These basic properties include, but are not limited to: mass, center of mass position, inertia tensor, and geometric dimensions.

[0154] Step S40b: Determine the joint data of the legged robot.

[0155] The joint data includes, but is not limited to: type, degrees of freedom, angle / displacement, force / torque parameters, motion speed, acceleration, transmission parameters, motor parameters, etc.

[0156] Step S41: Schedule the preset dynamic equations.

[0157] The dynamic equations include, but are not limited to: Lagrange equations, Newton-Euler equations, Archimedes-Euler equations, tangent stiffness matrix method, rigid body dynamics model, finite element method, example dynamics model, incremental mechanics method, multibody system dynamics, etc.

[0158] Step S42: Determine the actuator.

[0159] The actuator includes, but is not limited to, the torso, hand, leg, and foot.

[0160] Step S43: Construct the dynamic model file.

[0161] Based on the above examples, this application also provides a possible implementation of a legged robot test model construction and integrated testing device, specifically... Figure 8 This is a schematic diagram of the architecture of a legged robot test model construction and integrated testing device provided in an embodiment of the present invention. (See attached diagram.) Figure 8 The device 50 includes: a data acquisition terminal 500, a scene modeling configuration module 501, a sensor modeling configuration module 502, a dynamics modeling configuration module 503, and an integrated testing module 504. The data acquisition terminal 500 and the scene modeling configuration module 501 are used to generate scene model files, which are used to define the virtual scene information required for legged robot testing.

[0162] The sensor modeling configuration module 502 is used to generate a sensor simulation model file coupled with the environment based on the scene model file. The sensor simulation model file is used to define the perception function information of the legged robot.

[0163] The dynamic modeling configuration module 503 is used to combine the basic information of the legged robot with the characteristics of the scene to generate a verified dynamic model file, in which the dynamic model file is used to define the motion function information of the legged robot.

[0164] The integrated test module 504 is used to load scene model files, sensor simulation model files, and dynamic model files to complete integrated collaborative simulation testing; and to update the configuration of the test process.

[0165] Optionally, the data acquisition terminal 500 is specifically used to collect multi-dimensional data after selection and calibration, and generate standard data files through noise reduction, deduplication and format conversion.

[0166] The scene modeling configuration module 501 is specifically used to verify standard data files, call built-in tools to complete 3D reconstruction and topology optimization, and generate a scene 3D model; identify the scene 3D model and separate dynamic and static elements, add unique identifiers and calibrate spatial positions, and generate a decoupled scene element factor set, where dynamic elements are dynamic objects in the scene 3D model; static elements are static objects in the scene 3D model; generate data labels; each data label has a mapping relationship with dynamic or static elements; integrate the scene element factor set, export a model file in a universally compatible format, and generate a model attribute report containing the number of elements and parameter ranges.

[0167] Optionally, the sensor modeling configuration module 502 is specifically used to select from a set of multi-dimensional sensing sensors, call the sensor-specific configuration template that matches the scene model file; define mandatory and optional parameters; bind the mandatory and optional parameters with the environmental parameters in the scene model file to generate a sensor parameter set; adapt the physics engine to the sensor parameter set; import the parameters in the sensor parameter set to generate a sensor simulation model; and establish a coupling relationship between the sensor simulation model and the environmental parameters in the scene model file to form a sensor simulation model file.

[0168] Optionally, the integrated testing module 504 is specifically used to load the scene model file, sensor simulation model file, and dynamics model file; perform compatibility verification on the scene model file, sensor simulation model file, and dynamics model file to obtain a multi-dimensional model set that passes verification; configure the transmission protocol for the multi-dimensional model set; test the interaction link between the multi-dimensional model set and the robot according to the transmission protocol, and construct the data interaction rule set corresponding to the multi-dimensional model set; initialize the scene environment parameters and the motion state of dynamic elements according to the environmental parameters in the scene model file, start the scene environment simulation and update it in real time to generate a real-time simulated virtual test scene; collect scene environment data and robot motion data in real time in the virtual test scene according to the sensor simulation model file, and transmit them to the dynamics model file according to the protocol and frequency of the data interaction rule set; verify that the dynamics model file retrieves the preset combination of dynamic equations, calculates motion commands according to the collected sensor data, and drives the robot simulation model to perform obstacle-crossing actions; record all test data in real time and generate an integrated test report.

[0169] Optionally, Figure 9 This is a schematic structural block diagram of an electronic device provided in an embodiment of this application. Please refer to... Figure 9The electronic device 60 includes a memory 601, a processor 602, and a communication interface 603. The memory 601, processor 602, and communication interface 603 are electrically connected directly or indirectly to achieve data transmission or interaction. For example, these components can be electrically connected to each other through one or more communication buses or signal lines. The memory 601 can be used to store software programs and modules, such as the program instructions / modules corresponding to the legged robot test model construction and integration test device 500 provided in this application embodiment. The processor 602 executes various functional applications and data processing by executing the software programs and modules stored in the memory 601. The communication interface 603 can be used for signaling or data communication with other node devices.

[0170] The memory 601 may be, but is not limited to, random access memory (RAM), read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), etc.

[0171] Processor 602 can be an integrated circuit chip with signal processing capabilities. Processor 602 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0172] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can also be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0173] In addition, the functional modules in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0174] If a function is implemented as a software module and sold or used as an independent program product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This program product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0175] The above are merely preferred embodiments of this application and are not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0176] It will be apparent to those skilled in the art that this application is not limited to the details of the exemplary embodiments described above, and that this application can be implemented in other specific forms without departing from the spirit or essential characteristics of this application. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of this application is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within this application. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for constructing and integrating a test model for a legged robot, characterized in that, The method includes: The steps for generating scene model files by combining the data acquisition terminal with the scene modeling configuration module include: After the data acquisition terminal is selected and calibrated, it collects multi-dimensional data, and generates standard data files through noise reduction, deduplication and format conversion. The scene modeling configuration module verifies the standard data file, calls built-in tools to complete 3D reconstruction and topology optimization, and generates a 3D model of the scene. The scene modeling configuration module identifies the scene 3D model and separates dynamic and static elements. After adding a unique identifier, it calibrates the spatial position and generates a decoupled scene element factor set. The dynamic elements are dynamic objects in the scene 3D model, and the static elements are static objects in the scene 3D model. The scene modeling configuration module generates data tags; each data tag has a mapping relationship with the dynamic element or the static element. The scene modeling configuration module integrates the scene element factor set, exports a model file in a universally compatible format, and generates a model attribute report containing the number of elements and parameter ranges. The scene model file is used to define the virtual scene information required for testing the legged robot. The sensor modeling configuration module generates a sensor simulation model file coupled with the environment based on the scene model file, wherein the sensor simulation model file is used to define the perception function information of the legged robot; The dynamics modeling configuration module combines the basic information of the legged robot with the scene characteristics to generate a verified dynamics model file, wherein the dynamics model file is used to define the motion function information of the legged robot; The integrated testing module loads the scene model file, the sensor simulation model file, and the dynamics model file; The integrated testing module performs compatibility verification on the scene model file, the sensor simulation model file, and the dynamics model file to obtain a multi-dimensional model set that passes the verification. The integrated testing module configures the transmission protocol for the multidimensional model set; The integration testing module performs interaction link testing on the multidimensional model set according to the transmission protocol, and constructs a data interaction rule set corresponding to the multidimensional model set. The integrated testing module initializes the scene environment parameters and the motion state of the dynamic elements according to the environment parameters in the scene model file, starts scene environment simulation and updates it in real time, and generates a real-time simulated virtual test scene. The integrated testing module collects scene environment data and robot motion data in real time in the virtual test scenario according to the sensor simulation model file, and transmits them to the dynamic model file according to the protocol and frequency of the data interaction rule set; The integrated testing module verifies that the dynamic model file retrieves a preset combination of dynamic equations, calculates motion commands based on the collected sensor data, and drives the robot simulation model to perform obstacle-crossing actions. The integration test module records all test data in real time and generates an integration test report. Update the test process configuration.

2. The method as described in claim 1, characterized in that, The step of the sensor modeling configuration module generating a sensor simulation model file coupled with the environment based on the scene model file includes: The sensor modeling and configuration module selects from a set of multi-dimensional sensing sensors and calls up sensor-specific configuration templates that match the scene model file; The sensor modeling and configuration module determines the required and optional parameters; The sensor modeling configuration module binds the required parameters, the optional parameters, and the environmental parameters in the scene model file to generate a sensor parameter set; The sensor modeling and configuration module adapts the physics engine to the sensor parameter set; The sensor modeling configuration module imports parameters from the sensor parameter set to generate a sensor simulation model; The sensor modeling configuration module establishes a coupling relationship between the sensor simulation model and the environmental parameters in the scene model file to form the sensor simulation model file.

3. A test model construction and integrated testing device for a legged robot, characterized in that, The data acquisition terminal and the scene modeling configuration module are used to generate scene model files. The scene model files are used to define the virtual scene information required for legged robot testing. Specifically, after selection and calibration, multi-dimensional data is collected and processed through noise reduction, deduplication, and format conversion to generate standard data files. The scene modeling configuration module is specifically used to verify the standard data file, call built-in tools to complete 3D reconstruction and topology optimization, and generate a 3D scene model; identify the 3D scene model and separate dynamic and static elements, add unique identifiers and calibrate spatial positions, and generate a decoupled scene element factor set, wherein the dynamic elements are dynamic objects in the 3D scene model; the static elements are static objects in the 3D scene model; generate data tags; each data tag has a mapping relationship with the dynamic or static elements; integrate the scene element factor set, export a model file in a universally compatible format, and generate a model attribute report containing the number of elements and parameter ranges; The sensor modeling configuration module is used to generate a sensor simulation model file coupled with the environment based on the scene model file, wherein the sensor simulation model file is used to define the perception function information of the legged robot; The dynamic modeling configuration module is used to combine the basic information of the legged robot with the scene characteristics to generate a verified dynamic model file, wherein the dynamic model file is used to define the motion function information of the legged robot; The integrated testing module is specifically used to: load the scene model file, the sensor simulation model file, and the dynamics model file; perform compatibility verification on the scene model file, the sensor simulation model file, and the dynamics model file to obtain a multidimensional model set that passes verification; configure a transmission protocol for the multidimensional model set; perform interaction link testing on the multidimensional model set according to the transmission protocol, and construct a data interaction rule set corresponding to the multidimensional model set; initialize scene environment parameters and the motion state of the dynamic elements according to the environmental parameters in the scene model file, start scene environment simulation and update it in real time to generate a real-time simulated virtual test scene; collect scene environment data and robot motion data in real time in the virtual test scene according to the sensor simulation model file, and transmit them to the dynamics model file according to the protocol and frequency of the data interaction rule set; verify that the dynamics model file retrieves a preset combination of dynamic equations, calculates motion commands according to the collected sensor data, and drives the robot simulation model to perform obstacle-crossing actions; record all test data in real time and generate an integrated test report; and configure and update the test process.

4. The apparatus as described in claim 3, characterized in that, The sensor modeling and configuration module is specifically used to select from a set of multi-dimensional sensing sensors, call the sensor-specific configuration template that matches the scene model file; determine the required parameters and optional parameters; bind the required parameters and optional parameters with the environmental parameters in the scene model file to generate a sensor parameter set; adapt the physics engine to the sensor parameter set; import the parameters in the sensor parameter set to generate a sensor simulation model; and establish a coupling relationship between the sensor simulation model and the environmental parameters in the scene model file to form the sensor simulation model file.

5. An electronic device, characterized in that, include: Memory, used to store one or more programs; processor; When one or more programs are executed by a processor, the method of any one of claims 1-2 is implemented.

6. A program product, characterized in that, When the program product contains a program that is executed by a processor, it implements the method as described in any one of claims 1-2.

Images