Method and device for reconstructing a scenario of abnormal operation of a robot, electronic device and storage medium

Abstract

A robot abnormal operation scene reconstruction method, device, electronic device and computer readable storage medium, wherein the robot abnormal operation scene reconstruction method comprises: when the robot abnormally operates, acquiring first scene reconstruction data including pose data, sensor data and obstacle map of the robot from the robot, determining a timing moment of the robot abnormal operation, selecting second scene reconstruction data in the first scene reconstruction data, the second scene reconstruction data including pose data, sensor data and obstacle map of the robot generated after the timing moment, and reconstructing a scene where the robot abnormally operates according to a preset scene reconstruction algorithm and the second scene reconstruction data to obtain a scene graph containing information of the robot abnormal operation, which can improve the efficiency of solving the robot operation abnormality and positioning problem.

Description

Technical Field

[0001] This application relates to the field of robotics, and in particular to a method, apparatus, electronic device, and computer-readable storage medium for reconstructing a scene of abnormal robot operation. Background Technology

[0002] In real-world applications, the path planning function of mobile robots sometimes leads to malfunctions. These malfunctions are highly dependent on the specific scenario, and the unpredictable nature of these scenarios makes them difficult to reproduce. Therefore, how to reproduce the scenario in which malfunctions occur has become a major technical problem that needs to be solved.

[0003] In existing technologies, log replay is typically used for testing to locate problems and verify the effectiveness of modifications. However, log replay may fail to reproduce the problem scene due to data errors, reducing the efficiency of finding the cause of operational anomalies. Summary of the Invention

[0004] This application provides a method, apparatus, electronic device, and computer-readable storage medium for reconstructing a scene of abnormal robot operation, which can solve the problems of abnormal robot operation and positioning.

[0005] This application provides a method for reconstructing a scene of abnormal robot operation, including:

[0006] When the robot malfunctions, first scene reconstruction data is obtained from the robot. The first scene reconstruction data includes the robot's pose data, sensor data, and obstacle map.

[0007] The timing of the abnormal robot operation is determined, and the timing is the time before the time of the abnormal robot operation. Second scene reconstruction data is selected from the first scene reconstruction data. The second scene reconstruction data includes the robot's pose data, sensor data, and obstacle map generated after the timing. The scene in which the abnormal robot operation occurred is reconstructed according to the preset scene reconstruction algorithm and the second scene reconstruction data to obtain a scene map containing information about the abnormal robot operation.

[0008] This application also provides a scene reconstruction device for robot malfunction, comprising: an acquisition module for acquiring first scene reconstruction data from the robot when the robot malfunctions, the first scene reconstruction data including the robot's pose data, sensor data, and obstacle map; a determination module for determining the timing of the robot's malfunction, the timing being a time before the time of the robot's malfunction; a selection module for selecting second scene reconstruction data from the first scene reconstruction data, the second scene reconstruction data including the robot's pose data, sensor data, and obstacle map generated after the timing; and a reconstruction module for reconstructing the scene where the robot malfunctions according to a preset scene reconstruction algorithm and the second scene reconstruction data, to obtain a scene map containing information about the robot's malfunction.

[0009] One aspect of this application also provides an electronic device, including:

[0010] The system includes a memory and a processor; the memory stores executable program code; the processor, coupled to the memory, calls the executable program code stored in the memory to execute the scene reconstruction method for abnormal robot operation as described above.

[0011] In one aspect, this application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the scene reconstruction method for abnormal robot operation as described above.

[0012] As can be seen from the above embodiments of this application, when the robot malfunctions, the first scene reconstruction data generated by the task in which the malfunction occurred is obtained, including data such as the robot's pose, sensors, and obstacle maps. The timing of the malfunction is determined, and the robot's pose, sensors, and obstacle maps generated after that timing in the first scene reconstruction data are used as the second scene reconstruction data. Based on the preset scene reconstruction algorithm and the second scene reconstruction data, the scene of the robot's malfunction is reconstructed, and the scene at that time is restored to obtain a scene map containing information about the robot's malfunction at that time. The cause of the robot's malfunction can be intuitively determined. This scene map can serve as a basis for solving robot localization problems and improve the efficiency of solving robot malfunctions and localization. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 A flowchart illustrating a method for reconstructing a scene of abnormal robot operation according to an embodiment of this application;

[0015] Figure 2 A flowchart illustrating a method for reconstructing a scene of abnormal robot operation, provided in another embodiment of this application;

[0016] Figure 3 A schematic diagram of the structure of a scene reconstruction device for abnormal robot operation provided in an embodiment of this application;

[0017] Figure 4 A schematic diagram of the structure of a scene reconstruction device for abnormal robot operation provided in another embodiment of this application.

[0018] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0019] 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 the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] This application provides a method for reconstructing a scene in case of abnormal robot operation. Specifically, when the robot performs a movement-based task, it records the following data in real time: robot pose data, sensor data, obstacle map, running path, and running status data. The recorded data is marked with time information, specifically timestamps, with the start time of this relative time being the start time of the robot's current task. Specifically, the data with time information is used to generate a log file in chronological order of occurrence and stored in the robot's memory.

[0021] When the robot moves, abnormal walking behavior may occur due to path planning issues, sudden obstacle appearances, or problems with the robot's sensors or controllers. The robot may be unable to move normally within the current scene, such as circling in a narrow area, moving back and forth, or swaying left and right, and may be unable to continue moving forward. This is considered abnormal robot operation. When an abnormal robot operation occurs, the terminal remotely retrieves the log file from the robot and obtains the first scene reconstruction data from the log file. This first scene reconstruction data includes the robot's pose data, sensor data, and obstacle map; this first scene reconstruction data is generated during the task in which the abnormal operation occurred.

[0022] Furthermore, by replaying logs or reporting abnormal information from on-site personnel, the exact moment when the robot malfunctioned is identified. From this first scene reconstruction data, the robot's pose data, sensor data, and obstacle map generated after that moment are selected, and the scene at that time is reconstructed using a pre-set scene reconstruction algorithm, such as the Gazebo simulator.

[0023] The terminal can be a computer, mobile phone, or wearable smart device, and typically has a structure including an arithmetic unit, controller, memory, input devices, and output devices. The following describes in detail the scene reconstruction method for abnormal robot operation.

[0024] See Figure 1 This application provides a flowchart illustrating a method for reconstructing a scene of abnormal robot operation according to an embodiment. This method can be applied to terminals, such as... Figure 1 As shown, the method specifically includes:

[0025] S101. When the robot malfunctions, obtain the first scene reconstruction data from the robot.

[0026] Specifically, the first scene reconstruction data obtained when an abnormal operation occurs is the data of the current task that is experiencing the abnormal operation; data for tasks that are not experiencing abnormal operations need not be obtained.

[0027] The first scene reconstruction data includes the data required by the scene reconstruction algorithm of this embodiment. Corresponding to the scene reconstruction algorithm, the scene reconstruction algorithm preferably uses the Gazebo simulator, but can also be a simulator such as Vrep, CoppeliaSim, Stage, Webots, etc.

[0028] The first scene reconstruction data includes the robot's pose data, sensor data, and obstacle map;

[0029] The pose data includes the robot's position and direction of movement.

[0030] Sensors include inertial measurement units (composed of multiple accelerometers and gyroscopes), atmospheric meters, compasses, positioning systems, camera devices, sonar sensors, and lidar, etc.

[0031] An obstacle map is a grid map that represents obstacle information. This obstacle map is updated in real time. Specifically, it generates a map composed of grids, and based on sensor information, obstacles appearing within an 8×8 grid area are projected onto the map.

[0032] The first scene reconstruction data is data saved in real time during robot operation. Log files are generated according to the chronological order of the data and stored in the robot's memory. When generating the log file, unique time information is marked for the data. Specifically, timestamps can be added to the log file. These timestamps can be timestamps corresponding to absolute time, which is the current Greenwich Mean Time or Beijing Time; or timestamps corresponding to relative time, where the start time of the relative time is the start time of the robot's current task. The time in the robot system is calculated based on the time difference with this start time.

[0033] Specifically, in this embodiment, the log file generated by the robot's task when the abnormal operation occurred is obtained. The log file includes the robot's pose data, sensor data, and obstacle map.

[0034] S102. Determine the timing of any abnormal robot operation;

[0035] This timing point is the starting point for collecting data on abnormal robot operation, an earlier time than the actual occurrence of the abnormal operation, and preceded by a preset duration, such as 3 seconds. To prevent data causing the abnormal operation from being generated before this time, the acquisition time of the data needed for scene reconstruction is moved forward by this preset duration. This allows for the acquisition of more data for scene reconstruction, increasing the success rate of identifying the cause of the abnormal operation in the reconstructed scene graph.

[0036] Specifically, you can check the log playback to see when the robot malfunctioned, or confirm the abnormal time recorded by the user as the time when the robot malfunctioned, and then subtract the preset duration from that time to get the timing. For example, if the time is 8:00 AM on the same day and the preset duration is 2 seconds, then the timing will be 7:58 AM on the same day.

[0037] S103. Select second scene reconstruction data from the first scene reconstruction data. The second scene reconstruction data includes robot pose data, sensor data and obstacle map generated after the timing time.

[0038] The data required for scene reconstruction, including pose data, sensor data, and obstacle maps generated at and after the timing point, are selected from the first scene reconstruction data and used as the data basis for the scene reconstruction algorithm.

[0039] S104. Based on the preset scene reconstruction algorithm and the second scene reconstruction data, reconstruct the scene where the robot malfunctions, and obtain a scene map containing information about the robot's malfunction.

[0040] The scene reconstruction algorithm reconstructs the scene of the robot's abnormal operation based on the pose data, sensor data, and obstacle map generated at the time of the incident and thereafter. The reconstructed scene map contains information about the robot's abnormal operation, including its position, orientation, and obstacle information. This scene map is then output from the terminal's display interface, allowing users to intuitively see the robot's abnormal operation and determine the cause of the abnormal operation.

[0041] In this embodiment, when the robot malfunctions, the system acquires first scene reconstruction data generated by the task that caused the malfunction, including data such as the robot's pose, sensor data, and obstacle map. The system determines the timing of the malfunction and uses the robot's pose, sensor data, and obstacle map generated after that timing in the first scene reconstruction data as second scene reconstruction data. Based on a preset scene reconstruction algorithm and the second scene reconstruction data, the scene of the robot's malfunction is reconstructed, restoring the scene at that time and obtaining a scene map containing information about the robot's malfunction at that time. This allows for a direct assessment of the cause of the robot's malfunction, and the scene map can serve as a basis for solving robot localization problems, improving the efficiency of resolving robot malfunctions and localization.

[0042] See Figure 2 The following is a flowchart illustrating the implementation of a method for reconstructing a robot's abnormal operation scene according to another embodiment of the present invention. This method can be applied to terminals, such as... Figure 2 As shown, the method specifically includes:

[0043] S201. When the robot malfunctions, obtain the first scene reconstruction data from the log file of the task in which the robot malfunctioned.

[0044] The first scene reconstruction data includes the robot's pose data, sensor data, and obstacle map;

[0045] The robot records pose data, sensor data, obstacle map, running path, and running status data in real time. The running path includes the robot's planned path and the path it has already traveled. The planned path includes global paths and local paths. The running status data includes the robot's real-time running form, running scene, and running task data. The running form refers to the robot's normal and abnormal movement patterns. Abnormal patterns include spinning in place, short back-and-forth movements, and short swaying movements, which affect the robot's normal movement on the running path. The running scene refers to the environment in which the task is performed, such as a restaurant or hotel. The running task refers to the task that the robot needs to perform, such as food delivery or goods transportation.

[0046] All the above data is saved in chronological order as the robot's log file.

[0047] S202. Determine the timing of any abnormal robot operation;

[0048] By replaying the logs, the time when the robot malfunctioned can be confirmed based on the robot's pose data, sensor data, obstacle map, running path, and running status data in the log file. Alternatively, the time when the robot malfunctioned can be obtained from the abnormal information reported by the user, and the time when the robot malfunctioned can be subtracted from the time when the robot malfunctioned can be obtained as the timing time.

[0049] S203. Select second scene reconstruction data from the first scene reconstruction data. The second scene reconstruction data includes robot pose data, sensor data and obstacle map generated after the timing time.

[0050] The data required for scene reconstruction, including pose data, sensor data, and obstacle maps generated at and after the timing point, are selected from the first scene reconstruction data and used as input data for the scene reconstruction algorithm.

[0051] S204. Based on the preset scene reconstruction algorithm and the second scene reconstruction data, reconstruct the scene where the robot malfunctions, and obtain a scene map containing information about the robot's malfunction.

[0052] Using the pre-set Gazebo simulator, and the robot's pose data, sensor data, and obstacle map generated after the timing point, a scene map of the robot's abnormal operation is reconstructed, which displays information about the robot's abnormal operation.

[0053] Sensor data may include: linear acceleration and angular velocity provided by the inertial measurement unit; robot height provided by the atmospheric sensor; orientation data provided by the compass; longitude, latitude, and altitude data provided by the positioning system; images of the front view from the camera; obstacle detection data provided by the sonar sensor; ranging data from the lidar; depth data provided by the depth camera, etc.

[0054] Pose data, sensor data, and obstacle maps include, but are not limited to, the descriptions in the above embodiments. The specific calculations are subject to the calculation requirements of the Gazebo simulator. For the specific calculation principles and processes of the Gazebo simulator, please refer to relevant common knowledge and tutorial materials, which will not be elaborated here.

[0055] S205. In the reconstructed scene, the modified robot path planning algorithm is invoked to simulate the robot's operation, and the modified robot path planning algorithm is verified based on the robot's operating status.

[0056] After the scene reconstruction is completed, the abnormal operation status of the robot in the scene is analyzed to find the cause of the abnormal operation. Based on the cause, the current path planning algorithm of the robot is corrected so that the robot will no longer have the above-mentioned abnormal operation status when using the corrected path planning algorithm.

[0057] Using a pre-set scene reconstruction algorithm, based on the robot's real-time recorded pose data, sensor data, obstacle map, running path and running status data, as well as the corrected path planning algorithm, the scene is reproduced. The robot then uses the corrected robot path planning algorithm to replan its path. If the abnormal running state is not resolved, the path planning algorithm is corrected again, and the scene reconstruction algorithm is used again to reproduce the scene to see if the robot still exhibits abnormal running states. This process continues until the robot no longer exhibits abnormal running states, at which point the path planning algorithm correction is successful.

[0058] In this context, the starting point for the robot's operation in the reconstructed scene is the position the robot is at at the time of timing. The corrected robot path planning algorithm is obtained by modifying the current path planning algorithm based on the reasons for the robot's abnormal operation analyzed in the reconstructed scene.

[0059] It can also analyze the data in the acquired robot log files to construct scenarios where this type of robot is prone to problems but has not yet encountered any issues. For example, it can construct scenarios with more complex obstacle locations, quantities, and task content than existing scenarios. The preset scenario reconstruction algorithm is used to obtain a simulation scenario of this scenario. If the robot malfunctions in the simulation scenario, the robot's path planning algorithm is corrected based on the information of the malfunction. The scenario reconstruction algorithm is then used to verify whether the corrected path planning algorithm has solved the problem of robot malfunction. If it has solved the problem, the algorithm correction is successful. If it has not solved the problem, the path planning algorithm is corrected again and verified again until the algorithm correction is successful.

[0060] The aforementioned method of using scene reconstruction algorithms to correct and verify the robot's path planning algorithm on the terminal, without needing to write the corrected algorithm into the actual robot and run it in a real-world scenario, can greatly improve the efficiency of solving robot path planning problems. Furthermore, while abnormal robot operating states are related to the specific scenario, they may not necessarily occur in actual operation. However, if the corrected path planning algorithm has been verified successfully on the terminal in that scenario, then using the corrected path planning algorithm on the actual site will not result in abnormal operating states, thus ensuring the robot's normal operation more efficiently.

[0061] It should be noted that robot path planning algorithms include not only path planning itself, but also sub-algorithms for obstacle avoidance, stabilization, scheduling, and other aspects. In other words, a robot can achieve its entire operation through path planning algorithms. In specific scenarios, a particular sub-algorithm within the path planning algorithm can be modified; this is not a limitation here.

[0062] S206. Save the corrected data on abnormal robot operation.

[0063] Specifically, the scene diagram containing information about the robot's abnormal operation is saved, as well as the robot's pose data, sensor data, obstacle map, running path, and running status data in the log file corresponding to the scene diagram after the timing of the robot's abnormal operation.

[0064] The above data will be saved as historical cases and retrieved as a reference for correcting robot path planning algorithms when needed for future scenario replication.

[0065] For technical details of each of the above steps, please refer to the aforementioned... Figure 2 The description of the illustrated embodiment will not be repeated here.

[0066] It should be noted that, in another embodiment, after determining the timing of the robot's abnormal operation, there are two ways to obtain information about the robot's abnormal operation: one is through scene reconstruction in steps S204 to S205; the other is through log playback.

[0067] By replaying the logs, we can analyze the robot's pose data, sensor data, obstacle map, running path, and running status data in the log files after the timing of the abnormal robot operation. This information includes the robot's position information, orientation information, obstacle information, etc. when the abnormal robot is running.

[0068] If log playback fails to reveal information about abnormal robot operation, proceed to step S203. Other steps are the same as... Figure 2 The embodiments shown are the same.

[0069] In this embodiment, when the robot malfunctions, first scene reconstruction data generated by the task in which the malfunction occurred is acquired, including data such as the robot's pose, sensor data, and obstacle map. The timing of the malfunction is determined, and the robot's pose, sensor data, and obstacle map data generated after that timing moment in the first scene reconstruction data are used as second scene reconstruction data. Based on a preset scene reconstruction algorithm and the second scene reconstruction data, the scene of the robot's malfunction is reconstructed, restoring the scene at that time and obtaining a scene map containing information about the robot's malfunction at that time. This allows for a direct assessment of the cause of the robot's malfunction. This scene map can serve as a basis for solving robot localization problems, improving the efficiency of resolving robot malfunctions and localization. Furthermore, when problems arise during reconstruction, the scene reconstruction algorithm is corrected. The corrected algorithm is verified, and related data is saved, further improving the efficiency of resolving robot malfunctions and localization, and optimizing the scene reconstruction algorithm.

[0070] See Figure 3 This application provides a schematic diagram of a scene reconstruction device for abnormal robot operation according to an embodiment. For ease of explanation, only the parts relevant to this embodiment are shown. This device can be installed in a terminal. The device includes:

[0071] The acquisition module 301 is used to acquire first scene reconstruction data from the robot when the robot is running abnormally. The first scene reconstruction data includes the robot's pose data, sensor data and obstacle map.

[0072] The determination module 302 is used to determine the timing of the abnormal operation of the robot, and the timing is the time before the time of the abnormal operation of the robot;

[0073] The selection module 303 is used to select the second scene reconstruction data from the first scene reconstruction data. The second scene reconstruction data includes the robot's pose data, sensor data, and obstacle map generated after the timing time.

[0074] The reconstruction module 304 is used to reconstruct the scene where the robot malfunctions based on the preset scene reconstruction algorithm and the second scene reconstruction data, and obtain a scene map containing information about the robot's malfunction.

[0075] The technical details of this embodiment are described in the foregoing embodiments and will not be repeated here.

[0076] In this embodiment, when the robot malfunctions, the system acquires first scene reconstruction data generated by the task that caused the malfunction, including data such as the robot's pose, sensor data, and obstacle map. The system determines the timing of the malfunction and uses the robot's pose, sensor data, and obstacle map generated after that timing in the first scene reconstruction data as second scene reconstruction data. Based on a preset scene reconstruction algorithm and the second scene reconstruction data, the scene of the robot's malfunction is reconstructed, restoring the scene at that time and obtaining a scene map containing information about the robot's malfunction at that time. This allows for a direct assessment of the cause of the robot's malfunction, and the scene map can serve as a basis for solving robot localization problems, improving the efficiency of resolving robot malfunctions and localization.

[0077] See Figure 4 This application provides a schematic diagram of a scene reconstruction device for robot malfunction according to an embodiment. For ease of explanation, only the parts relevant to this embodiment are shown. This device can be installed in a terminal. This device is relative to... Figure 3 The illustrated embodiment differs in the following ways:

[0078] The acquisition module 301 is also used to acquire first scene reconstruction data from the log file of the task in which the robot malfunctions when the robot is running abnormally. The log file also includes the robot's real-time recorded running path and running status data. The running status data includes the robot's real-time running form, running scene and running task data.

[0079] The determination module 302 is also used to confirm the time when the robot malfunctions by playing back the logs, based on the robot's pose data, sensor data, obstacle map, running path and running status data in the log file, or to obtain the time when the robot malfunctions by obtaining the time from the abnormal information reported by the user; and to subtract a preset duration from the time to obtain the timing time.

[0080] The device also includes an analysis module 401, which is used to analyze the robot's pose data, sensor data, obstacle map, running path and running status data in the log file after the timing of the robot's abnormal operation through log playback, so as to obtain information about the robot's abnormal operation.

[0081] If the information about the robot's abnormal operation cannot be obtained through log playback, the reconstruction module 304 is triggered to perform the step of reconstructing the scene where the robot's abnormal operation occurred based on the preset scene reconstruction algorithm and the second scene reconstruction data, so as to obtain a scene map containing the information about the robot's abnormal operation.

[0082] Furthermore, the reconstruction module 304 is also used to reconstruct the scene map where the robot malfunctions by utilizing the preset Gazebo simulator, as well as the robot's pose data, sensor data, and obstacle map generated after the timing point. The scene map displays information about the robot's malfunction.

[0083] The device also includes: a correction module 402, used to call a corrected robot path planning algorithm to simulate the operation of the robot in the reconstructed scene, and to verify the corrected robot path planning algorithm according to the operation status of the robot;

[0084] Wherein, the starting point for the robot's operation in the reconstructed scene is the position of the robot at the time of the timing, and the corrected robot path planning algorithm is an algorithm obtained by correcting the current path planning algorithm of the robot based on the reasons for the abnormal operation of the robot analyzed in the reconstructed scene.

[0085] The storage module 403 is used to store a scene diagram containing information about abnormal robot operation, as well as robot pose data, sensor data, obstacle map, running path and running status data in the log file corresponding to the scene diagram after the timing of the abnormal robot operation.

[0086] The technical details of this embodiment are described in the foregoing embodiments and will not be repeated here.

[0087] In this embodiment, when the robot malfunctions, first scene reconstruction data generated by the task that caused the malfunction is acquired, including data such as the robot's pose, sensor data, and obstacle map. The timing of the malfunction is determined, and the robot's pose, sensor data, and obstacle map data generated after that timing in the first scene reconstruction data are used as second scene reconstruction data. Based on a preset scene reconstruction algorithm and the second scene reconstruction data, the scene of the robot's malfunction is reconstructed, restoring the scene at that time and obtaining a scene map containing information about the robot's malfunction at that time. This allows for a direct assessment of the cause of the robot's malfunction. This scene map can serve as a basis for solving robot localization problems, improving the efficiency of resolving robot malfunctions and localization. When problems arise during reconstruction, the scene reconstruction algorithm is corrected, and the corrected algorithm is verified and the relevant data is saved, further improving the efficiency of resolving robot malfunctions and localization, and optimizing the scene reconstruction algorithm.

[0088] like Figure 4 As shown, this application embodiment also provides an electronic device, including a memory 100 and a processor 200. The processor 200 may be the processing module 304 in the scene reconstruction device for abnormal robot operation in the above embodiment. The memory 100 may be, for example, a hard disk drive memory, a non-volatile memory (e.g., flash memory or other electronically programmable and erasable memory used to form a solid-state drive), a volatile memory (e.g., static or dynamic random access memory), etc., and this application embodiment does not impose any limitations.

[0089] The memory 100 stores executable program code; the processor 200, coupled to the memory 100, calls the executable program code stored in the memory to execute the scene reconstruction method for abnormal robot operation as described above.

[0090] Furthermore, embodiments of the present invention also provide a computer-readable storage medium, which may be disposed in the terminal of the above embodiments, and the computer-readable storage medium may be as described above. Figure 4 The memory 100 in the illustrated embodiment. A computer program is stored on this computer-readable storage medium, which, when executed by a processor, implements the aforementioned... Figure 1 and Figure 2 The illustrated embodiment describes a method for reconstructing a scene of abnormal robot operation. Furthermore, the computer storage medium can also be any medium capable of storing program code, such as a USB flash drive, external hard drive, read-only memory (ROM), RAM, magnetic disk, or optical disk.

[0091] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0092] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0093] The above is a description of the method, apparatus, electronic device, and computer-readable storage medium for reconstructing abnormal robot operation scenarios provided by the present invention. For those skilled in the art, based on the ideas of the embodiments of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for reconstructing a scene of abnormal robot operation, characterized in that, include: When the robot malfunctions, the first scene reconstruction data is obtained from the log file of the task in which the robot malfunctioned. The log file also includes the robot's real-time recorded running path and running status data. The running status data includes the robot's real-time running form, running scene, and running task data. The first scene reconstruction data includes the robot's pose data, sensor data, and obstacle map. By replaying the logs, the moment when the robot malfunctioned is confirmed based on the robot's pose data, sensor data, obstacle map, running path, and running status data in the log file; or, the moment when the robot malfunctioned is obtained from the abnormal information reported by the user; the moment is subtracted from the time to obtain the timing time, which is the time before the moment when the robot malfunctioned. Select second scene reconstruction data from the first scene reconstruction data. The second scene reconstruction data includes the robot's pose data, sensor data, and obstacle map generated after the timing time. Based on the preset scene reconstruction algorithm and the second scene reconstruction data, the scene in which the robot malfunctions is reconstructed to obtain a scene map containing information about the robot's malfunction.

2. The method according to claim 1, characterized in that, After determining the timing of the abnormal robot operation, the process also includes: By replaying the logs, the robot's pose data, sensor data, obstacle map, running path, and running status data in the log file after the timing of the abnormal robot operation are analyzed to obtain information about the abnormal robot operation. If the information about the robot's abnormal operation cannot be obtained through the log playback, then the step of reconstructing the scene where the robot's abnormal operation occurred based on the preset scene reconstruction algorithm and the second scene reconstruction data is executed to obtain a scene map containing the information about the robot's abnormal operation.

3. The method according to any one of claims 1 to 2, characterized in that, The step of reconstructing the scene where the robot malfunctions, based on a preset scene reconstruction algorithm and the second scene reconstruction data, to obtain a scene map containing information about the robot's malfunction includes: Using a pre-set Gazebo simulator, and the robot's pose data, sensor data, and obstacle map generated after the timing point, a scene map showing the abnormal robot operation is reconstructed, and the scene map displays information about the abnormal robot operation.

4. The method according to claim 1, characterized in that, The method further includes: In the reconstructed scenario, the modified robot path planning algorithm is invoked to simulate the robot's operation, and the modified robot path planning algorithm is verified based on the robot's operating status. Wherein, the starting point for the robot's operation in the reconstructed scene is the position of the robot at the time of the timing, and the corrected robot path planning algorithm is an algorithm obtained by correcting the current path planning algorithm of the robot based on the reasons for the abnormal operation of the robot obtained from the analysis in the reconstructed scene.

5. The method according to claim 4, characterized in that, The method further includes: Save the scene diagram containing information about the abnormal operation of the robot, and save the robot's pose data, sensor data, obstacle map, running path and running status data in the log file corresponding to the scene diagram after the timing of the abnormal operation of the robot.

6. A scene reconstruction device for abnormal robot operation, characterized in that, include: The acquisition module is used to acquire first scene reconstruction data from the log file of the task in which the robot malfunctions when the robot is running abnormally. The log file also includes the robot's real-time recorded running path and running status data. The running status data includes the robot's real-time running form, running scene and running task data. The first scene reconstruction data includes the robot's pose data, sensor data and obstacle map. The determination module is used to determine the time when the robot malfunctioned by playing back the logs, based on the robot's pose data, sensor data, obstacle map, running path, and running status data in the log file, or to obtain the time when the robot malfunctioned by the user from the abnormal information reported by the user; and to obtain the timing time by subtracting a preset duration from the timing time, wherein the timing time is the time before the time when the robot malfunctioned. The selection module is used to select second scene reconstruction data from the first scene reconstruction data. The second scene reconstruction data includes the robot's pose data, sensor data, and obstacle map generated after the timing time. The reconstruction module is used to reconstruct the scene where the robot malfunctions, based on a preset scene reconstruction algorithm and the second scene reconstruction data, to obtain a scene map containing information about the robot's malfunction.

7. An electronic device, characterized in that, include: Memory and processor; The memory stores executable program code; The processor coupled to the memory calls the executable program code stored in the memory to execute the scene reconstruction method for abnormal robot operation as described in any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the scene reconstruction method for abnormal robot operation as described in any one of claims 1 to 5.

Images