An isaacsim-based simulation method for a humanoid robot in a steel chemical testing laboratory

Abstract

The application discloses a steel detection and chemical analysis laboratory humanoid robot simulation method based on IsaacSim, relates to the technical field of humanoid robot simulation methods, and comprises the following steps: step 1, fine modeling of the environmental elements of a steel detection and chemical analysis laboratory; step 2, collection of a humanoid robot model in an IsaacSim simulation environment; step 3, configuration of the work flow of different tasks of the robot in the simulation scene; step 4, collection and analysis of the data of the robot when performing a task in a virtual environment; and step 5, feedback adjustment of an algorithm or control parameters based on the simulation results to improve the adaptability and stability of the robot in a complex laboratory environment. Through the method, a user can comprehensively verify an algorithm in a simulation environment, including a control algorithm, a visual recognition algorithm, a motion planning algorithm and a series of key algorithms of the humanoid robot, so that the stability and effectiveness of the humanoid robot in an actual environment are ensured.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of humanoid robot simulation methods, in particular to a steel inspection and chemical testing laboratory humanoid robot simulation method based on IsaacSim. BACKGROUND

[0002] Robots can replace manual experiments in steel inspection and chemical testing laboratories to avoid accidental risks that may occur during manual inspection. Robots applied in actual scenes usually need to be verified multiple times to ensure the possibility of experiments in actual scenes.

[0003] The defects of the existing robot simulation method are:

[0004] 1. The existing technology KR101475210B1 discloses an APPRARATUS for simulating robots. This technology does not have a method for simulating key algorithms of humanoid robots in a virtual environment, and cannot feed back and improve algorithms in a simulation environment. Therefore, a steel inspection and chemical testing laboratory humanoid robot simulation method based on IsaacSim is needed to solve the problem, which can simulate the feedback and improvement of the behavior of the robot in the simulation environment.

[0005] 2. The existing technology US5366896A discloses a laboratory system operated by a robot. This technology does not have a technology for providing virtual simulation of robots. When a robot performs an experiment in a laboratory, it cannot avoid potential risks during the experiment through a virtual environment. When a robot simulation fails, it is easy to cause actual loss of goods. Therefore, a steel inspection and chemical testing laboratory humanoid robot simulation method based on IsaacSim is needed to solve the problem, which can avoid potential risks to the robot and the surrounding environment through virtual simulation.

[0006] 3. The existing technology KR1020130122288A discloses a robot three-dimensional simulation method and system. This technology does not have a method for providing behavior testing of robots in a variety of simulated task environments. Robots cannot test behavior in multiple scenarios. Therefore, a steel inspection and chemical testing laboratory humanoid robot simulation method based on IsaacSim is needed to solve the problem, which can simulate robot behavior simulation in a variety of different task environments.

[0007] 4、Prior art CN115801869B discloses a distributed architecture-based inspection robot simulation system and method, which cannot complete the debugging simulation of the robot through the simulation method. The robot needs to debug the accessories of the robot multiple times during actual debugging, increasing the debugging time. Therefore, an IsaacSim-based steel inspection laboratory humanoid robot simulation method for debugging the accessories of the robot through the simulation method is needed to solve the problem. SUMMARY

[0008] One object of the present application is to provide an IsaacSim-based steel inspection laboratory humanoid robot simulation method that can solve the technical problems in the prior art.

[0009] To achieve the above object, the present application provides the following technical solution: an IsaacSim-based steel inspection laboratory humanoid robot simulation method, which comprises the following steps:

[0010] Step 1: Fine modeling of environmental elements in the steel inspection laboratory, fine modeling of each key component and environmental element in the steel inspection laboratory, establishing a digital model consistent with the actual environment, and deploying the model to the database of the IsaacSim simulation environment to ensure that the simulation scene accurately reflects the operating environment and spatial layout of the real laboratory;

[0011] Step 2: Integrate the humanoid robot model in the IsaacSim simulation environment, load the control system including the vision recognition, motion planning and control algorithm modules, make it execute specific tasks in the simulation scene, set the perception sensors and motion drivers for the robot model, the perception sensors include cameras and laser radars, make it have consistent environmental perception and motion ability with the real robot;

[0012] Step 3: Configure the workflow of different tasks of the robot in the simulation scene, set the typical task flow of the humanoid robot in the simulation scene, including sampling, analysis, classification and handling tasks in the inspection laboratory, execute the task sequence in the simulation, obtain the performance data of the algorithm in the experimental task, verify and optimize the control accuracy and task completion efficiency of the robot;

[0013] Step 4: Collect and analyze the data of the robot executing tasks in the virtual environment, collect and analyze the performance of the humanoid robot in different task scenarios during the simulation process, including vision recognition accuracy, path planning time efficiency, operation stability, collision detection parameters and overall evaluation of algorithm performance, and generate an experimental report;

[0014] Step 5: Based on the simulation results, feedback adjustment algorithm or control parameters to improve the adaptability and stability of the robot in complex laboratory environment.

[0015] Preferably, step 1 creates a virtual environment of a steel inspection laboratory on the IsaacSim simulation platform, which includes equipment and tools in the laboratory, and by configuring the corresponding physical properties of the equipment and tools in the laboratory, ensures realistic physical interaction between the equipment and the humanoid robot in the virtual environment.

[0016] Device layout: According to the actual device layout of the steel inspection laboratory, set the specific position of the equipment in the simulation environment, to ensure that the simulation environment is consistent with the real laboratory;

[0017] Fine modeling of tools: For the tools used by the robot in the experiment, fine modeling and function configuration are carried out to enable the robot to operate effectively in the simulation.

[0018] Preferably, the humanoid robot model in step 2 includes kinematics and dynamics model, sensor simulation and physical characteristics of the robot, the kinematics and dynamics model includes its skeleton, joints, actuators and sensors, to ensure that the model imitates the joint movement of human, and can flexibly perform tasks such as grabbing, carrying and operating instruments, each joint is simulated using high-precision physical engine to ensure that the behavior of the robot in the simulation environment is consistent with the actual hardware.

[0019] Physical characteristics of the robot: Set the physical properties of the robot, including mass, center of gravity and inertia, so that the robot can exhibit realistic motion performance in the virtual environment, simulate the load and dynamic behavior in actual work.

[0020] Preferably, the sensor simulation integrates vision sensor and IMU sensor to capture information in the virtual environment and feed back to the control system, helping the robot to perform precise operations in the simulation.

[0021] Preferably, step 2 includes the construction of NVIDIA Omnigraph, task flow design and feedback control, and the construction of NVIDIA Omnigraph: Omnigraph is used for communication between virtual environment and external ROS system, through Omnigraph, the robot can interact with external systems in the virtual environment.

[0022] Task flow design: The tasks of the humanoid robot in the steel inspection laboratory are designed by the robot control algorithm, including grabbing samples, pouring reagents, operating heating equipment, etc.

[0023] Feedback Control: Through Omnigraph, Isaac Sim publishes sensor feedback data, allowing the robot to adjust its actions in real-time, dynamically optimizing its operation.

[0024] Preferably, the humanoid robot model in step 2 also includes ROS2 communication integration, bidirectional communication, and control interface. ROS2 communication integration simulates the interaction between the real robot and external systems, using ROS2 as the communication middleware between the robot and the simulation environment. By integrating it with IsaacSim, consistency between the simulation and the real system is ensured.

[0025] Bidirectional communication: Through ROS2, the robot receives task instructions from the control system and feeds back its internal state and sensor data to the external system, fully simulating the behavior of real robots communicating with the outside world during simulation.

[0026] Control interface: ROS2 provides a standardized control interface for the robot, ensuring that the action control, data collection, and feedback mechanisms in the simulation are consistent with the real robot. After simulation, the ROS2 communication part is applied to the actual robot deployment, reducing the debugging time in the actual environment.

[0027] Preferably, step 3 designs the task flow of the robot in the steel inspection and chemical testing process based on the virtual environment and the robot model, including the following key operation steps:

[0028] Sample sampling: The robot grabs and takes out the steel sample from the sample container and places it into the designated testing equipment;

[0029] Reagent addition: The robot accurately pours the required chemical reagent from the reagent bottle and adds it to the sample for chemical reaction;

[0030] Equipment operation: The robot operates the experimental equipment by simulating hand movements;

[0031] Result reading: The robot reads the result data displayed by the experimental equipment through the built-in sensor and transmits the data back to the control center for processing through the simulation software;

[0032] These operations are simulated in real-time through the physical engine of IsaacSim to ensure the accuracy of the robot's actions and adaptability to the laboratory environment.

[0033] Preferably, the different task scenarios in step 4 include normal workflow and abnormal scenario testing. Normal workflow: simulates the standard operation process in the steel inspection and chemical testing laboratory, ensuring that the robot completes the whole process from sample handling to data reading.

[0034] Abnormal scenario testing: when a sudden event is involved in the simulation, verify whether the robot can respond appropriately in the event of a sudden situation and avoid further loss.

[0035] Preferably, the different task scenarios in step 4 also include extreme working condition testing: simulate complex or extreme environmental conditions to test the operation performance and stability of the robot in these environments.

[0036] Preferably, in step 5, during the simulation run, the system records each step of the robot's operation, including motion path, operation time, task success rate and equipment collision, and through analysis of these data, identifies the operation bottlenecks and potential risk points of the robot, and the specific analysis methods include: path optimization: analyze the motion trajectory of the robot when performing tasks, optimize its joint motion and path planning, reduce redundant actions, and improve work efficiency;

[0037] Collision detection: record the collision of the robot with the environment or equipment, adjust the action planning and operation accuracy, and avoid equipment damage or robot injury in actual operation;

[0038] According to the simulation data, iteratively optimize the robot action control algorithm and task flow to ensure that it can safely and efficiently complete the test tasks in the real laboratory environment.

[0039] Compared with the prior art, the beneficial effects of the present application are:

[0040] 1、Through the method, the user can perform comprehensive algorithm verification in the simulation environment, including a series of key algorithms such as the control algorithm, visual recognition algorithm and motion planning algorithm of the humanoid robot, to ensure its stability and effectiveness in the actual environment.

[0041] 2、The present application effectively avoids potential risks that the robot may cause to itself and the surrounding environment, and provides a safe and controllable test platform, thereby optimizing the operation efficiency and task completion quality of the robot and promoting the wide application of humanoid robots in the steel metallurgical industry test and assay scene.

[0042] 3、The present application simulates the performance of the humanoid robot in different task scenarios, including normal workflow, abnormal scenario testing and abnormal scenario testing, which can test the operation performance and stability of the robot in these environments, and reduce the simulation cost of the robot in the actual scene.

[0043] 4、The present application makes the behavior and control of the humanoid robot simple through the simulation of communication and feedback, and reduces the debugging time of the humanoid robot in the actual environment. BRIEF DESCRIPTION OF DRAWINGS

[0044] Figure 1The steel inspection and chemical analysis laboratory humanoid robot simulation method flow chart of the present application. DETAILED DESCRIPTION

[0045] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0046] Please refer to Figure 1 The IsaacSim-based steel inspection and chemical analysis laboratory humanoid robot simulation method includes the following steps:

[0047] Step 1: Fine modeling of environmental elements in the steel inspection and chemical analysis laboratory, fine modeling of each key component and environmental element in the steel inspection and chemical analysis laboratory, establishing a digital model consistent with the actual environment, and deploying the model to the database of the IsaacSim simulation environment to ensure that the simulation scene accurately reflects the operating environment and spatial layout of the real laboratory;

[0048] Step 2: Set up a humanoid robot model in the IsaacSim simulation environment, load the control system, including vision recognition, motion planning and control algorithm modules, so that it can perform specific tasks in the simulation scene, set up perception sensors and motion drivers for the robot model, including cameras and laser radars, so that it has consistent environmental perception and motion capabilities with real robots;

[0049] Step 3: Configure the workflow of different tasks of the robot in the simulation scene, set up the typical task flow of the humanoid robot in the simulation scene, including sampling, analysis, classification and handling tasks in the inspection and chemical analysis laboratory, execute the task sequence in the simulation to obtain the performance data of the algorithm in the experimental task, verify and optimize the control accuracy and task completion efficiency of the robot;

[0050] Step 4: Collect and analyze the data of the robot when performing tasks in the virtual environment, collect and analyze the performance of the humanoid robot in different task scenarios in the simulation process, including vision recognition accuracy, path planning time efficiency, operation stability, collision detection parameters and overall evaluation of algorithm performance, and generate an experimental report;

[0051] Step 5: Adjust the algorithm or control parameters based on the simulation results to improve the adaptability and stability of the robot in complex laboratory environments.

[0052] Step 1: Create a virtual environment for the steel inspection laboratory in Isaac Sim simulation platform, including the equipment and tools in the laboratory, and configure the corresponding physical properties of the equipment and tools in the laboratory to ensure realistic physical interaction between the equipment and the humanoid robot in the virtual environment.

[0053] Device layout: According to the actual equipment layout of the steel inspection laboratory, set the specific location of the equipment in the simulation environment to ensure consistency between the simulation environment and the real laboratory.

[0054] Fine modeling of tools: For the tools used by the robot in the experiment, fine modeling and function configuration are performed to enable the robot to operate effectively in the simulation.

[0055] The humanoid robot model in Step 2 includes kinematics and dynamics models, sensor simulation, and robot physical properties, including its skeleton, joints, actuators, and sensors, to ensure that the model mimics human joint movement and can flexibly perform tasks such as grabbing, carrying, and operating complex instruments for laboratory tests. Each joint is simulated using a high-precision physics engine to ensure that the robot's behavior in the simulation environment is consistent with the actual hardware.

[0056] Robot physical properties: Set the physical properties of the robot, including mass, center of gravity, and inertia, to enable the robot to exhibit realistic motion performance in the virtual environment and simulate the load and dynamic behavior in actual work.

[0057] Sensor simulation integrates vision sensors and IMU sensors to capture information in the virtual environment and feed back to the control system, helping the robot perform precise operations in the simulation.

[0058] Step 2 includes the construction of NVIDIA Omnigraph, task flow design, and feedback control. Build NVIDIA Omnigraph: Omnigraph is used for communication between the virtual environment and the external ROS environment, allowing the virtual robot to interact with external systems (such as ROS) through Omnigraph.

[0059] Task flow design: The logic of the robot's task execution is handled by the robot control algorithm, including grabbing samples, pouring reagents, and operating heating equipment. Omnigraph serves as a communication bridge to help the robot obtain task instructions and data from the external environment in the virtual environment and pass these instructions to the robot for execution.

[0060] Feedback Control: Through Omnigraph, Isaacsim publishes sensor information in the virtual environment to the external environment, including IMU information, image information, and robot state information to the control algorithm, combining feedback information from sensors, the control algorithm adjusts the robot's actions in real time to ensure the continuity and accuracy of the actions, allowing the robot control algorithm to obtain specific environmental information and robot state information.

[0061] The humanoid robot model in step 2 also includes ROS2 communication integration, bidirectional communication, and control interface. ROS2 communication integration: simulates the interaction between the real robot and the external system, uses ROS2 as the communication middleware between the robot and the simulation environment, and integrates it with IsaacSim to ensure consistency between the simulation and the real system;

[0062] Bidirectional communication: through ROS2, the robot receives task instructions from the control system and feeds back its internal state and sensor data to the external system, fully simulating the behavior of the real robot communicating with the outside world during simulation;

[0063] Control interface: ROS2 provides a standardized control interface for the robot, ensuring that the action control, data collection, and feedback mechanisms in the simulation are consistent with the real robot. After the simulation is complete, the ROS2 communication part is applied to the actual robot deployment, reducing the debugging time in the actual environment.

[0064] Step 3, based on the virtual environment and robot model, designs the task flow of the robot in the steel chemical testing process, including the following key operation steps:

[0065] Sample sampling: the robot grabs and takes out the steel sample from the sample container and places it into the designated testing equipment;

[0066] Reagent addition: the robot accurately pours the required chemical reagent from the reagent bottle and adds it to the sample for chemical reaction;

[0067] Equipment operation: the robot operates the experimental equipment by simulating hand movements;

[0068] Result reading: the robot reads the result data displayed by the experimental equipment through the built-in sensor and transmits the data back to the control center for processing through the simulation software;

[0069] These operations are simulated in real time through the physical engine of IsaacSim to ensure the accuracy of the robot's actions and adaptability to the laboratory environment.

[0070] Different task scenarios in Step 4 include normal workflow and abnormal scenario testing. Normal workflow: simulate the standard operation process in the steel inspection laboratory, ensure that the robot completes the whole process from sample processing to data reading;

[0071] Abnormal scenario testing: when emergencies are involved in simulation, verify whether the robot can respond appropriately in emergency situations and avoid further loss.

[0072] Different task scenarios in Step 4 also include extreme condition testing: simulate complex or extreme environmental conditions to test the operation performance and stability of the robot in these environments.

[0073] In Step 5, during the simulation run, the system records each step of the robot's operation, including motion path, operation time, task success rate and equipment collision, through analysis of these data, identify the operation bottlenecks and potential risk points of the robot, specific analysis methods include: path optimization: analyze the motion trajectory of the robot when executing tasks, optimize its joint motion and path planning, reduce redundant actions, and improve work efficiency;

[0074] Collision detection: record the collision between the robot and the environment or equipment, adjust the action planning and operation precision, avoid equipment damage or robot injury in actual operation;

[0075] According to the simulation data, iteratively optimize the robot action control algorithm and task flow to ensure that it can safely and efficiently complete the test tasks in the real laboratory environment.

[0076] Please refer to Figure 1 , the present application provides an embodiment;

[0077] The steel inspection laboratory humanoid robot simulation method based on IsaacSim includes the following steps:

[0078] Step 1: Fine modeling of environmental elements in the steel inspection laboratory, fine modeling of each key component and environmental element in the steel inspection laboratory, establishing a digital model consistent with the actual environment, and deploying the model to the database of IsaacSim simulation environment to ensure that the simulation scene accurately reflects the operation environment and spatial layout of the real laboratory;

[0079] Create a virtual environment of the steel inspection laboratory on the IsaacSim simulation platform, which includes equipment and tools in the laboratory, such as experimental tables, chemical analysis instruments, sample processing equipment, heaters, cooling devices, etc., by configuring corresponding materials, friction coefficients and gravity and other physical properties for the equipment and tools in the laboratory, to ensure that the equipment in the virtual environment and the humanoid robot produce realistic physical interaction;

[0080] Device Layout: According to the actual steel chemical laboratory equipment layout, set the specific location of the device in the simulation environment, ensure that the simulation environment is consistent with the real laboratory;

[0081] Fine modeling of tools: For tools used by robots in experiments, such as samplers, beakers, stirrers, etc., fine modeling and function configuration are carried out to enable the robot to operate effectively in simulation.

[0082] Step 2: Integrate humanoid robot model in IsaacSim simulation environment, load control system including vision recognition, motion planning and control algorithm modules, make it execute specific tasks in simulation scene, set perception sensors and motion drivers for robot model, perception sensors include cameras and lidar, make it have consistent environment perception and motion ability with real robot;

[0083] Humanoid robot model in step 2 includes kinematics and dynamics model, sensor simulation and robot physical characteristics, kinematics and dynamics model includes its skeleton, joints, actuators and sensors, etc., to ensure that the model imitates human joint movement and flexibly performs complex tasks such as grabbing, carrying and operating instruments, each joint uses high-precision physical engine for simulation to ensure that the behavior of the robot in the simulation environment is consistent with the actual hardware.

[0084] Robot physical characteristics: Set the physical properties of the robot, including mass, center of gravity and inertia, etc., so that the robot exhibits realistic motion performance in the virtual environment, simulating the load and dynamic behavior in actual work.

[0085] Sensor simulation integrates multiple sensor modules, including vision sensors and IMU sensors, etc., to capture information in the virtual environment and feed back to the control system, helping the robot perform precise operations in simulation.

[0086] Humanoid robot model also includes the construction of NVIDIA Omnigraph, task flow design and feedback control, build NVIDIA Omnigraph: Omnigraph is used for communication between virtual environment and external ROS environment, through Omnigraph, virtual robot interacts with external systems (such as ROS) for data exchange;

[0087] Task flow design: The logic of robot task execution is responsible for robot control algorithm, including grabbing samples, pouring reagents and operating heating equipment, Omnigraph as a communication bridge, helps the robot to get task instructions and data from the external environment in the virtual environment, and passes these instructions to the robot for execution;

[0088] Feedback Control: Through Omnigraph, Isaacsim publishes sensor information in the virtual environment to the external environment, including IMU information, image information, and robot state information to the control algorithm, combining the feedback information of the sensors, the control algorithm adjusts the robot's actions in real time to ensure the coherence and accuracy of the actions, allowing the robot control algorithm to obtain specific environmental information and robot state information.

[0089] The humanoid robot model also includes ROS2 communication integration, bidirectional communication, and control interface. ROS2 communication integration simulates the interaction between the real robot and the external system, using ROS2 as the communication middleware between the robot and the simulation environment. By integrating it with IsaacSim, consistency between the simulation and the real system is ensured.

[0090] Bidirectional communication: Through ROS2, the robot receives task instructions from the control system and feeds back its internal state and sensor data to the external system, fully simulating the behavior of the real robot communicating with the outside world during simulation.

[0091] Control interface: ROS2 provides a standardized control interface for the robot, ensuring that the motion control, data acquisition, and feedback mechanisms in the simulation are consistent with the real robot. After simulation is complete, the ROS2 communication part is applied to the actual robot deployment, reducing the debugging time in the actual environment.

[0092] Step 3: Configure the robot's workflow for different tasks in the simulation scene. Set up the typical task flow for the humanoid robot in the simulation scene, including sampling, analysis, classification, and handling tasks in the chemical laboratory. By executing the task sequence in the simulation, obtain the performance data of the algorithm in the experimental task, and verify and optimize the control accuracy and task completion efficiency of the robot.

[0093] Step 3 designs the task flow of the robot in the steel chemical testing process based on the virtual environment and the robot model, including the following key operation steps:

[0094] Sample sampling: The robot grabs and takes out the steel sample from the sample container and places it into the designated testing equipment.

[0095] Reagent addition: The robot accurately pours the required chemical reagent from the reagent bottle and adds it to the sample for chemical reaction.

[0096] Equipment operation: The robot simulates hand movements to operate experimental equipment, such as starting the heater, controlling the mixer, and setting temperature parameters.

[0097] Result reading: The robot reads the result data displayed by the experimental equipment through the built-in sensor and transmits the data back to the control center for processing through the simulation software.

[0098] These operations are simulated in real-time by the physics engine of IsaacSim to ensure the accuracy of the robot's movements and adaptability to the laboratory environment.

[0099] Step 4: Collect and analyze data on the robot's performance in the virtual environment, including visual recognition accuracy, path planning efficiency, operation stability, collision detection parameters, and overall algorithm performance, and generate an experimental report.

[0100] The different task scenarios in Step 4 include normal workflow and abnormal scenario testing. Normal workflow: simulate the standard operation process in a steel chemical testing laboratory to ensure that the robot completes the entire process from sample handling to data reading.

[0101] Abnormal scenario testing: involves unexpected events in the simulation, such as equipment failure, sample drop, reagent overflow, etc., to verify whether the robot can respond appropriately and avoid further damage in unexpected situations.

[0102] The different task scenarios in Step 4 also include extreme condition testing: simulate complex or extreme environmental conditions, such as high temperature or narrow laboratory space, to test the robot's operation performance and stability in these environments.

[0103] Step 5: Adjust the algorithm or control parameters based on the simulation results to improve the robot's adaptability and stability in complex laboratory environments.

[0104] In Step 5, the system records each step of the robot's operation during simulation, including motion path, operation time, task success rate, and equipment collision, etc. Through analysis of these data, the operation bottlenecks and potential risk points of the robot are identified. Specific analysis methods include: path optimization: analyze the robot's motion trajectory when performing tasks, optimize its joint motion and path planning, reduce unnecessary actions, and improve work efficiency;

[0105] Collision detection: record the collision between the robot and the environment or equipment, adjust the action planning and operation precision, and avoid equipment damage or robot injury in actual operation;

[0106] According to the simulation data, iterative optimization is carried out to adjust the robot's action control algorithm and task process, ensuring that it can safely and efficiently complete the testing tasks in the real laboratory environment.

[0107] It is apparent for a person skilled in the art that the present application is not limited to the details of the above described exemplary embodiments, but that it can be implemented in other concrete forms without departing from the spirit or essential characteristics of the application. Therefore, the embodiments should be considered in all respects as illustrative and not restrictive, the scope of the application being defined by the appended claims rather than the above description, and all changes coming within the meaning and equivalency range of the claims are intended to be embraced therein, no figure reference in the claims being considered limiting as to the scope of the claim concerned.

Claims

1. An IsaacSim-based simulation method for a humanoid robot in a steel chemical testing laboratory, characterized in that: The IsaacSim-based simulation method of the steel inspection laboratory humanoid robot comprises the following steps: Step 1: Fine modeling of environmental elements in the steel inspection laboratory, fine modeling of various key components and environmental elements in the steel inspection laboratory, establishing a digital model consistent with the actual environment, and deploying the model to the database of the IsaacSim simulation environment to ensure that the simulation scene accurately reflects the operating environment and spatial layout of the real laboratory; Step 2: In the IsaacSim simulation environment, integrate the humanoid robot model, load the control system, including the vision recognition, motion planning and control algorithm module, so that it can perform specific tasks in the simulation scene, set up perception sensors and motion drivers for the robot model, including cameras and laser radars, so that it has consistent environmental perception and motion ability with the real robot; Step 3: Configure the workflow of different tasks of the robot in the simulation scene, set the typical task flow of the humanoid robot in the simulation scene, including sampling, analysis, classification and handling tasks in the inspection laboratory, execute the task sequence in the simulation to obtain the performance data of the algorithm in the experimental task, verify and optimize the control accuracy and task completion efficiency of the robot; Step 4: Collect and analyze the data of the robot executing tasks in the virtual environment, collect and analyze the performance of the humanoid robot in different task scenarios in the simulation process, including vision recognition accuracy, path planning time efficiency, operation stability, collision detection parameters and comprehensive evaluation of algorithm performance, and generate an experimental report; Step 5: Based on the simulation results, adjust the algorithm or control parameters to improve the adaptability and stability of the robot in complex laboratory environments; The humanoid robot model in step 2 includes kinematics and dynamics models, sensor simulation and robot physical properties, kinematics and dynamics models include its skeleton, joints, actuators and sensors, to ensure that the model imitates the joint movement of humans, and can flexibly perform complex tasks such as grabbing, handling and operating instruments, each joint is simulated using a high-precision physics engine to ensure that the behavior of the robot in the simulation environment is consistent with the actual hardware; Robot physical properties: set the physical properties of the robot, including mass, center of gravity and inertia, so that the robot exhibits realistic motion performance in the virtual environment, simulating the load and dynamic behavior in actual work; The sensor simulation integrates vision sensors and IMU sensors to capture information in the virtual environment and feed back to the control system to help the robot perform precise operations in the simulation; The step 2 includes the construction of NVIDIA Omnigraph, task flow design and feedback control, and the construction of NVIDIA Omnigraph: Omnigraph is used for communication between the virtual environment and the external ROS environment, through Omnigraph, the virtual robot interacts with external systems for data exchange; Task flow design: The logic of the robot's task execution is handled by the robot control algorithm, including grabbing samples, pouring reagents, and operating heating equipment. Omnigraph serves as a communication bridge, helping the robot obtain task instructions and data from the external environment in the virtual environment. Feedback control: Through Omnigraph, Isaacsim publishes sensor information in the virtual environment to the external environment, including IMU information, image information, and robot state information to the control algorithm. Combined with the feedback information of the sensor, the control algorithm adjusts the robot's actions in real time to ensure the continuity and accuracy of the actions. The humanoid robot model in step 2 also includes ROS2 communication integration, bidirectional communication, and control interface. ROS2 communication integration simulates the interaction between the real robot and the external system, using ROS2 as the communication middleware between the robot and the simulation environment. By integrating it with IsaacSim, consistency between the simulation and the real system is ensured. Bidirectional communication: Through ROS2, the robot receives task instructions from the control system and feeds back its internal state and sensor data to the external system, fully simulating the behavior of the real robot communicating with the outside world during simulation. Control interface: ROS2 provides a standardized control interface for the robot, ensuring that the action control, data collection, and feedback mechanisms in the simulation are consistent with the real robot. After simulation, the ROS2 communication part is applied to the actual robot deployment, reducing the debugging time in the actual environment. Different task scenarios in step 4 include normal workflow and abnormal scenario testing. Normal workflow: simulates the standard operation process in the steel inspection laboratory to ensure that the robot completes the entire process from sample processing to data reading. Abnormal scenario testing: when a sudden event occurs in the simulation, it verifies whether the robot can respond appropriately and avoid further loss in the event of a sudden situation. Extreme condition testing: simulate complex or extreme environmental conditions to test the robot's operating performance and stability in these environments.

2. The Isaac Sim-based simulation method for a humanoid robot in a steel chemical testing laboratory according to claim 1, wherein: Step 1 creates a virtual environment of a steel inspection laboratory on the IsaacSim simulation platform, which includes the equipment and tools in the laboratory. By configuring the corresponding physical properties for the equipment and tools in the laboratory, realistic physical interactions between the equipment and the humanoid robot in the virtual environment are ensured. Device layout: according to the actual device layout of the steel inspection laboratory, the specific location of the equipment in the simulation environment is set to ensure consistency between the simulation environment and the real laboratory. Fine modeling of tools: for the tools used by the robot in the experiment, fine modeling and function configuration are performed to enable the robot to operate effectively in the simulation.

3. The Isaac Sim-based simulation method for a steel chemical testing laboratory humanoid robot according to claim 1, characterized in that: Step 3 designs the task flow of the robot in the steel inspection process based on the virtual environment and the robot model, including the following key operation steps: Sample sampling: the robot grabs and takes out the steel sample from the sample container and places it into the designated testing equipment; Reagent addition: the robot accurately pours the required chemical reagent from the reagent bottle and adds it to the sample for chemical reaction; Device operation: The robot operates the experimental equipment by simulating hand movements; Result reading: The robot reads the result data displayed by the experimental equipment through the built-in sensor and transmits the data back to the control center for processing through the simulation software; These operations are simulated in real time through the physical engine of IsaacSim to ensure the accuracy of the robot's movements and adaptability to the laboratory environment.

4. The Isaac Sim-based simulation method for a steel chemical testing laboratory humanoid robot according to claim 1, characterized in that: During the simulation process, the system records each step of the robot's operation, including the motion path, operation time, task success rate, and device collision situation. Through analysis of these data, the operation bottlenecks and potential risk points of the robot are identified. The specific analysis methods include: Path optimization: Analyze the motion trajectory of the robot when performing tasks, optimize its joint motion and path planning, reduce redundant actions, and improve work efficiency; Collision detection: Record the collision situation between the robot and the environment or equipment, adjust the action planning and operation accuracy, and avoid equipment damage or robot injury in actual operation; According to the simulation data, iterative optimization is carried out to adjust the robot action control algorithm and task process, ensuring that it can safely and efficiently complete the test tasks in the real laboratory environment.

Images