Robot location planning method and apparatus based on reinforcement learning
By employing a reinforcement learning-based robot position planning method, utilizing a positioning pen module and a deep neural network training model, the robot position planning is optimized, solving the applicability and efficiency issues of robot position planning and achieving high-precision industrial applications.
Patent Information
- Application Number
- PCT/CN2025/143696
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2024-12-20
- Filing Date
- 2025-12-18
- Publication Date
- 2026-06-25
AI Technical Summary
Existing technologies for robot location planning suffer from limitations in applicability, high levels of human intervention, and low efficiency.
A reinforcement learning-based robot position planning method is adopted. The object information is digitally described by a pre-set positioning pen module, an initial position planning model is trained using a deep neural network, and the scene is optimized based on the observation space and task reward logic to obtain an optimized robot position planning model.
It improves the accuracy and efficiency of robot positioning planning, meets the needs of various industrial scenarios, and significantly reduces the workload of manual operation.
Smart Images
Figure CN2025143696_25062026_PF_FP_ABST
Abstract
Description
Robot position planning method and device based on reinforcement learning
[0001] Cross-reference to Related Applications
[0002] The present application claims priority to the Beijing Xiaoyu Zhizao Technology Co., Ltd. filed on December 20, 2024, entitled "Robot position planning method and device based on reinforcement learning", Chinese patent application number "202411885997.3". TECHNICAL FIELD
[0003] The present application relates to the field of robot motion planning, and in particular to a robot position planning method and device based on reinforcement learning, an electronic device, and a storage medium. BACKGROUND
[0004] During the execution of a task by a robot, the robot as a whole or a part thereof often needs to move to a specified position to detect target positions, postures, and other information in the task using sensors. Among them, guiding the robot to detect at an appropriate position and posture and feeding back perception data that can meet the requirements of the target detection task is one of the key links in task execution. In related technologies, pre-recorded robot positions are used for fixed objects, which has great limitations in applicability, or manual control and guiding the robot arm to the corresponding position based on a teach pendant, which has high human involvement and high labor costs, and low efficiency. SUMMARY
[0005] The present application aims to at least partially solve one of the technical problems in the related art.
[0006] To this end, the first object of the present application is to propose a robot position planning method based on reinforcement learning, based on a trained robot position planning model, so that the robot planning accuracy meets the landing requirements of various industrial scenarios, and greatly reduces the workload of manual operation.
[0007] The second object of the present application is to propose a robot position planning device based on reinforcement learning.
[0008] The third object of the present application is to propose an electronic device.
[0009] The fourth object of the present application is to propose a non-transitory computer readable storage medium having computer instructions stored therein.
[0010] To achieve the above objects, the first aspect of the present application proposes a robot position planning method based on reinforcement learning, the method comprising:
[0011] The digital object data is obtained by presetting a positioning pen module to digitally describe position, shape and attitude information of an object to be controlled by the robot.
[0012] The digital object data is taken as input of a deep neural network of a simulation platform corresponding to the robot, and a control position of the object by the robot is taken as output to train an initial position planning model of the robot.
[0013] Based on an application scenario of the object to be controlled by the robot, an observation space and task reward logic corresponding to the initial position planning model are selected.
[0014] Based on the observation space and the task reward logic, the initial position planning model is optimized to obtain an optimized robot position planning model, which is deployed on the robot.
[0015] Based on the robot position planning model, the digital object data of the object to be controlled is predicted to obtain a target position of the object to be controlled by the robot.
[0016] To achieve the above purpose, a second aspect embodiment of the present application provides a robot position planning device based on reinforcement learning, which comprises:
[0017] A digital description module is configured to digitally describe position, shape and attitude information of an object to be controlled by the robot by a preset positioning pen module to obtain digital object data.
[0018] A training module is configured to take the digital object data as input of a deep neural network of a simulation platform corresponding to the robot, and take a control position of the object by the robot as output to train an initial position planning model of the robot.
[0019] A selection module is configured to select an observation space and task reward logic corresponding to the initial position planning model based on an application scenario of the object to be controlled by the robot.
[0020] An optimization module is configured to optimize the initial position planning model based on the observation space and the task reward logic to obtain an optimized robot position planning model, which is deployed on the robot.
[0021] A prediction module is configured to predict the digital object data of the object to be controlled based on the robot position planning model to obtain a target position of the object to be controlled by the robot.
[0022] To achieve the above objectives, a third aspect of the present invention provides an electronic device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the method described in the first aspect.
[0023] To achieve the above objectives, a fourth aspect of the present invention provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform the method described in the first aspect.
[0024] The robot position planning method, apparatus, electronic device, and storage medium based on reinforcement learning provided in this invention utilize a preset positioning pen module to digitize the position, shape, and posture information of the object to be manipulated by the robot. This digitized object data serves as the input to the deep neural network of the robot's corresponding simulation platform, while the robot's control position of the object is output, training an initial position planning model for the robot. Based on the object's application scenario, a corresponding observation space and task reward logic are selected to optimize the initial position planning model, resulting in a robot position planning model. This model then predicts the target position of the robot's manipulation of the digitized object data. Therefore, based on the trained robot position planning model, the robot's planning accuracy meets the requirements of various industrial scenarios, while significantly reducing the workload of manual operations.
[0025] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description
[0026] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:
[0027] Figure 1 is a flowchart illustrating a robot position planning method based on reinforcement learning provided in an embodiment of the present invention;
[0028] Figure 2 is a schematic diagram of a deep neural network structure provided in an embodiment of the present invention;
[0029] Figure 3 is an application flowchart of another robot position planning method based on reinforcement learning provided in an embodiment of the present invention;
[0030] Figure 4 is a schematic diagram of a robot position planning device based on reinforcement learning provided in an embodiment of the present invention. Detailed Implementation
[0031] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.
[0032] It should be noted that the acquisition, storage, use, and processing of data in the technical solution of this invention all comply with the relevant provisions of relevant laws and regulations.
[0033] The following description, with reference to the accompanying drawings, outlines a reinforcement learning-based robot position planning method, apparatus, electronic device, and storage medium according to embodiments of the present invention.
[0034] Figure 1 is a flowchart illustrating a robot position planning method based on reinforcement learning provided in an embodiment of the present invention.
[0035] As shown in Figure 1, the method includes the following steps:
[0036] Step 101: Using a preset positioning pen module, the position, shape, and posture information of the object to be manipulated by the robot are digitally described to obtain digital object data.
[0037] In some embodiments, the preset positioning pen module is a type of prompt module. A prompt module is a language template used to guide a large language model (LLM) to perform specific tasks or generate expected content. By providing explicit instructions or questions, the prompt helps users interact with the large model, reducing model uncertainty and thus improving the accuracy and structure of the output.
[0038] Specifically, the positioning pen module allows for very convenient operation to digitally describe the position, shape, and posture information of an object, which is then transmitted to the robot as a prompt for task planning.
[0039] The orientation information of the object includes, but is not limited to, horizontal, vertical, and tilted orientations; this embodiment does not specifically limit these orientations.
[0040] Step 102: The digitized object data is used as the input to the deep neural network of the robot's corresponding simulation platform, and the robot's control position of the object is used as the output to train the robot's initial position planning model.
[0041] In some embodiments, when the application scenario based on the object to be manipulated by the robot is weld inspection, and the robot is a robotic arm, the IsaacSim simulation platform is used to train a deep neural network. The deep neural network structure adopts a fully connected plus ReLU activation method, contains 7 hidden layers, and the number of neurons is [64, 64, 32, 32, 16, 8, 8].
[0042] Specifically, as shown in Figure 2, the number of neurons in the input layer and output layer of the deep neural network can both be 6 (,), and the number of neurons in the hidden layer ([64,64,32,32,16,8,8]) are as follows: , , , , , , .
[0043] Furthermore, in an application scenario where the robot needs to manipulate objects such as weld seam inspection, and the robot is a robotic arm, the position, shape, and orientation information of the weld seam, corresponding to the weld seam's digital data, are used as input to the deep neural network of the robot's simulation platform. The six-axis position of the robotic arm is used as the output to train the robot's initial position planning model. Thus, based on the IsaacSim simulation platform, the richness of the scenario and the concurrency of the learning and training process can be significantly and efficiently improved, and the cost is extremely low compared to the strong learning process in the actual environment.
[0044] When training deep neural networks using the IsaacSim simulation platform, the number of concurrent processes (training data) can be designed to be 4096.
[0045] Step 103: Based on the application scenario of the object that the robot needs to manipulate, select the observation space and task reward logic corresponding to the initial position planning model.
[0046] In some embodiments, where the application scenario based on the object to be manipulated by the robot is weld seam inspection, and the robot is a robotic arm, the observation space corresponding to the object includes weld seam information collected by the positioning pen, the current position of the robotic arm, and the joint information of the robotic arm. The task reporting logic includes the distance from the camera on the robotic arm to the weld seam, the angle between the camera and the base plate of the welded workpiece, whether the camera image fully covers the weld seam, collision penalty for the robotic arm, and abnormal action penalty. Therefore, based on rich application scenarios and large-scale parallel simulation training, the robot planning accuracy meets the needs of industrial implementation, and the planning efficiency far surpasses traditional solutions.
[0047] Step 104: Based on the observation space and task reward logic, the initial position planning model is optimized to obtain an optimized robot position planning model, which is then deployed on the robot.
[0048] In some embodiments, where the application scenario based on the robot's manipulated object is weld seam detection, and the robot is a robotic arm, the initial position planning model is optimized based on weld seam information, the robotic arm's current position, joint information, the distance from the camera on the robotic arm to the weld seam, the angle between the camera and the workpiece base plate, whether the camera image fully covers the weld seam, robotic arm collision penalties, and abnormal action penalties. This results in an optimized robot position planning model, which is then deployed on the robot. Therefore, in the stage of finding a specific weld seam in a welding mechanical wall, this method is not only applicable to weld seam detection in various non-standard scenarios with different positions and poses, but also significantly reduces the workload of manual operation.
[0049] Step 105: Based on the robot position planning model, predict the digitized object data of the object to be manipulated to obtain the target position for the robot to manipulate the object.
[0050] In some embodiments, taking weld inspection as an example, the position, shape and posture information of the weld to be welded can be obtained by a positioning pen. The corresponding real-time weld digital data is input into the robot position planning model, thereby predicting the target position of the robot arm when welding the weld and moving to the target position to complete the rapid and accurate welding of the weld without human intervention.
[0051] The robot position planning method based on reinforcement learning in this invention uses a pre-set positioning pen module to digitize the position, shape, and posture information of the object to be manipulated by the robot. This digitized object data serves as the input to the deep neural network of the robot's corresponding simulation platform, while the robot's control position of the object is the output, training an initial position planning model for the robot. Based on the object's application scenario, a corresponding observation space and task reward logic are selected to optimize the initial position planning model, resulting in a new robot position planning model. This model then predicts the target position of the robot's manipulation of the digitized object data. Therefore, based on the trained robot position planning model, the robot's planning accuracy meets the requirements of various industrial scenarios, while significantly reducing the workload of manual operations.
[0052] To clearly illustrate the previous embodiment, Figure 3 is an application flowchart of a robot position planning method based on reinforcement learning provided in this embodiment of the invention. Specifically, the position, shape, and posture information of the object to be manipulated by the robot can be digitally described through a preset prompt module (positioning pen module) to obtain digital object data; then, deep reinforcement learning task modeling is performed, which includes: defining the object's corresponding observation space, the deep neural network structure, the deep neural network output, and the deep neural network's task reward logic; then, simulation and deep reinforcement learning model training are performed, which includes simulation environment modeling (constructing the robot's initial position planning model based on the deep neural network structure and deep neural network output) and simulation training (simulating and training the neural network model of the robot's position based on the observation space and task reward logic). Thus, the neural network model is deployed as the robot position planning model on the robot (robotic arm) body.
[0053] To achieve the above embodiments, the present invention also proposes a robot position planning device based on reinforcement learning.
[0054] Figure 4 is a schematic diagram of a robot position planning device based on reinforcement learning provided in an embodiment of the present invention.
[0055] As shown in Figure 4, the reinforcement learning-based robot position planning device 40 includes: a digital description module 41, a training module 42, a selection module 43, an optimization module 44, and a prediction module 45.
[0056] The digital description module 41 is used to digitally describe the position, shape and posture information of the object to be manipulated by the robot through a preset positioning pen module, so as to obtain digital object data.
[0057] Training module 42 is used to take the digitized object data as input to the deep neural network of the robot's corresponding simulation platform and the robot's control position of the object as output to train the robot's initial position planning model.
[0058] Select module 43 to select the observation space and task reward logic corresponding to the initial position planning model based on the application scenario of the object to be manipulated by the robot.
[0059] The optimization module 44 is used to optimize the initial position planning model based on the observation space and task reward logic to obtain an optimized robot position planning model, and then deploy it on the robot.
[0060] The prediction module 45 is used to predict the digitized object data of the object to be manipulated based on the robot position planning model, so as to obtain the target position of the robot manipulating the object to be manipulated.
[0061] Furthermore, in one possible implementation of this invention, when the application scenario based on the object to be manipulated by the robot is weld seam detection, and the robot is a robotic arm, the observation space corresponding to the object includes weld seam information collected by the positioning pen, the current position of the robotic arm, and the joint information of the robotic arm. The task reporting logic includes the distance from the camera on the robotic arm to the weld seam, the angle between the camera and the base plate of the welded workpiece, whether the camera image fully covers the weld seam, robotic arm collision penalty, and abnormal action penalty.
[0062] Furthermore, in one possible implementation of this invention, the training module 42 is specifically used for:
[0063] In the application scenario where the robot needs to manipulate objects for weld inspection, and the robot is a robotic arm, the position, shape, and orientation information of the weld, corresponding to the weld digital data, are used as the input to the deep neural network of the robot's simulation platform, and the six-axis position of the robotic arm is used as the output, in order to train the robot's initial position planning model.
[0064] Furthermore, in one possible implementation of this invention, the optimization module 44 is specifically used for:
[0065] In the application scenario where the robot needs to manipulate objects for weld inspection, and the robot is a robotic arm, the initial position planning model is optimized based on weld information, the current position of the robotic arm, the joint information of the robotic arm, the distance from the camera on the robotic arm to the weld, the angle between the camera and the base plate of the welded workpiece, whether the camera image fully covers the weld, the robotic arm collision penalty, and the abnormal action penalty. This results in an optimized robot position planning model, which is then deployed on the robot.
[0066] Furthermore, in one possible implementation of the present invention, when the application scenario based on the object to be manipulated by the robot is weld seam inspection, and the robot is a robotic arm, the IsaacSim simulation platform is used for deep neural network training, and the deep neural network structure adopts a fully connected plus ReLU activation method, containing 7 hidden layer structures, and the number of neurons is [64,64,32,32,16,8,8].
[0067] It should be noted that the foregoing explanation of the method embodiments also applies to the apparatus of this embodiment, and will not be repeated here.
[0068] The robot position planning device based on reinforcement learning in this invention uses a preset positioning pen module to digitize the position, shape, and posture information of the object to be manipulated by the robot. This digitized object data serves as the input to the deep neural network of the robot's corresponding simulation platform, while the robot's control position of the object is the output, training an initial position planning model for the robot. Based on the object's application scenario, a corresponding observation space and task reward logic are selected to optimize the initial position planning model, resulting in a robot position planning model. This model then predicts the target position of the robot's manipulation of the digitized object data. Therefore, based on the trained robot position planning model, the robot's planning accuracy meets the requirements of various industrial scenarios, while significantly reducing the workload of manual operation.
[0069] To achieve the above embodiments, the present invention also proposes an electronic device, comprising:
[0070] At least one processor; and
[0071] A memory communicatively connected to the at least one processor; wherein,
[0072] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the aforementioned method.
[0073] To implement the above embodiments, the present invention also proposes a non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform the aforementioned method.
[0074] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.
[0075] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.
[0076] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of the invention pertain.
[0077] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.
[0078] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any of the following techniques known in the art, or a combination thereof: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.
[0079] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.
[0080] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.
[0081] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.
Claims
1. A robot position planning method based on reinforcement learning, characterized in that, The method includes: The robot uses a pre-set positioning pen module to digitally describe the position, shape, and posture of the object it needs to manipulate, in order to obtain digital object data. The digitized object data is used as the input to the deep neural network of the robot's corresponding simulation platform, and the robot controls the object's manipulation position as the output, so as to train the robot's initial position planning model. Based on the application scenario of the object that the robot needs to manipulate, the observation space and task reward logic corresponding to the initial position planning model are selected. Based on the observation space and task reward logic, the initial position planning model is optimized to obtain an optimized robot position planning model, which is then deployed on the robot. Based on the robot position planning model, the digitized object data of the object to be manipulated is predicted to obtain the target position for the robot to manipulate the object.
2. The method according to claim 1, characterized in that, In the application scenario where the robot needs to manipulate an object for weld inspection, and the robot is a robotic arm, the observation space corresponding to the object includes the weld information collected by the positioning pen, the current position of the robotic arm, and the joint information of the robotic arm. The task reporting logic includes the distance from the camera on the robotic arm to the weld, the angle between the camera and the base plate of the welded workpiece, whether the camera image fully covers the weld, the robotic arm collision penalty, and the abnormal action penalty.
3. The method according to claim 2, characterized in that, The step of using the digitized object data as input to the deep neural network of the robot's corresponding simulation platform, and using the robot's control position of the object as output, to train the robot's initial position planning model, includes: In the application scenario where the robot needs to manipulate objects for weld inspection, and the robot is a robotic arm, the position, shape, and orientation information of the weld, corresponding to the weld digital data, are used as the input to the deep neural network of the robot's simulation platform, and the six-axis position of the robotic arm is used as the output, in order to train the robot's initial position planning model.
4. The method according to claim 2, characterized in that, The process of optimizing the initial position planning model based on the observation space and task reward logic to obtain an optimized robot position planning model, and then deploying it on the robot, includes: In the application scenario where the robot needs to manipulate objects for weld inspection, and the robot is a robotic arm, the initial position planning model is optimized based on weld information, the current position of the robotic arm, the joint information of the robotic arm, the distance from the camera on the robotic arm to the weld, the angle between the camera and the base plate of the welded workpiece, whether the camera image fully covers the weld, the robotic arm collision penalty, and the abnormal action penalty. This results in an optimized robot position planning model, which is then deployed on the robot.
5. The method according to claim 2, characterized in that, In the application scenario where the robot needs to manipulate objects for weld inspection, and the robot is a robotic arm, the IsaacSim simulation platform is used to train a deep neural network. The deep neural network structure adopts a fully connected plus ReLU activation method, with 7 hidden layers and the number of neurons is [64,64,32,32,16,8,8].
6. A robot position planning device based on reinforcement learning, characterized in that, The device includes: The digital description module is used to digitally describe the position, shape, and posture information of the object to be manipulated by the robot through a preset positioning pen module, so as to obtain digital object data. The training module is used to take the digitized object data as the input to the deep neural network of the robot's corresponding simulation platform, and the robot's control position of the object as the output, so as to train the robot's initial position planning model. The selection module is used to select the observation space and task reward logic corresponding to the initial position planning model based on the application scenario of the object that the robot needs to manipulate. The optimization module is used to optimize the initial position planning model based on the observation space and task reward logic to obtain an optimized robot position planning model, which is then deployed on the robot. The prediction module is used to predict the target position of the robot to manipulate the object based on the digitized object data of the object to be manipulated by the robot position planning model.
7. The apparatus according to claim 6, characterized in that, In the application scenario where the robot needs to manipulate an object for weld inspection, and the robot is a robotic arm, the observation space corresponding to the object includes the weld information collected by the positioning pen, the current position of the robotic arm, and the joint information of the robotic arm. The task reporting logic includes the distance from the camera on the robotic arm to the weld, the angle between the camera and the base plate of the welded workpiece, whether the camera image fully covers the weld, the robotic arm collision penalty, and the abnormal action penalty.
8. The apparatus according to claim 7, characterized in that, The training module is specifically used for: In the application scenario where the robot needs to manipulate objects for weld inspection, and the robot is a robotic arm, the position, shape, and orientation information of the weld, corresponding to the weld digital data, are used as the input to the deep neural network of the robot's simulation platform, and the six-axis position of the robotic arm is used as the output, in order to train the robot's initial position planning model.
9. The apparatus according to claim 7, characterized in that, The optimization module is specifically used for: In the application scenario where the robot needs to manipulate objects for weld inspection, and the robot is a robotic arm, the initial position planning model is optimized based on weld information, the current position of the robotic arm, the joint information of the robotic arm, the distance from the camera on the robotic arm to the weld, the angle between the camera and the base plate of the welded workpiece, whether the camera image fully covers the weld, the robotic arm collision penalty, and the abnormal action penalty. This results in an optimized robot position planning model, which is then deployed on the robot.
10. The apparatus according to claim 7, characterized in that, In the application scenario where the robot needs to manipulate objects for weld inspection, and the robot is a robotic arm, the IsaacSim simulation platform is used to train a deep neural network. The deep neural network structure adopts a fully connected plus ReLU activation method, with 7 hidden layers and the number of neurons is [64,64,32,32,16,8,8].
11. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-5.
12. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-5.