A robot motion control method, device, equipment and storage medium

Abstract

The application discloses a robot motion control method and device, equipment and storage medium, relates to the computer technical field, and includes: obtaining the relative state information of a robot and a target mobile chassis, determining the target motion mode of the robot when getting on and off the chassis based on the current environment corresponding to the robot; if it is an auxiliary motion mode, generating a target action strategy of the robot relative to the target mobile chassis based on the relative state information and a compliant control method; if it is an autonomous motion mode, collecting demonstration action data of a human getting on and off a step or a chassis, and determining an initial action strategy by using an imitation learning algorithm, and optimizing the initial action strategy by reinforcement learning based on a preset reward function to obtain the target action strategy; the execution process of the target action strategy is monitored to adjust the target control parameter until the robot completes the getting on and off the chassis operation. The flexibility and adaptability of the robot in a complex scene are realized.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a robot motion control method, apparatus, device, and storage medium. Background Technology

[0002] As the integration of humanoid robots with wheeled / tracked mobile chassis increases, robots often need to complete the critical transition action of "getting on and off the chassis" in complex environments. For example, a robot may need to climb onto a mobile chassis from the ground, descend from the chassis to the ground, or frequently switch working modes between the chassis and the ground. This process involves changes in height, switching of support states, and human-robot or robot-robot collaboration, and has high requirements for stability and safety.

[0003] Existing solutions often treat "manually guided chassis mounting and dismounting" and "autonomous chassis mounting and dismounting" as separate issues, lacking a unified system architecture and multi-mode switching mechanism. They also fail to fully consider the coordinated adjustment between the chassis and the robot (such as position alignment and support method switching). Furthermore, existing technologies typically rely on manual remote control or preset action scripts, lacking learning-based adaptive control capabilities. This makes it difficult to adapt to mobile chassis at different heights or positions, thus limiting the expansion of practical application scenarios.

[0004] As can be seen from the above, how to achieve the flexibility and adaptability of robots in complex scenarios for mounting and dismounting chassis is an urgent problem to be solved. Summary of the Invention

[0005] In view of this, the purpose of this invention is to provide a robot motion control method, device, equipment, and storage medium, which enables the robot to achieve flexibility and adaptability in mounting and dismounting its chassis in complex scenarios. The specific solution is as follows: In a first aspect, this application provides a robot motion control method, including: Acquire the relative state information between the robot and the target mobile chassis; the relative state information includes the distance, relative angle, and height difference between the robot and the target mobile chassis; The target motion mode of the robot when it is on the upper and lower chassis is determined based on the current environment and the relative state information of the robot; the target motion mode includes assisted motion mode and autonomous motion mode. If the target motion mode is an auxiliary motion mode, then a target motion strategy for the robot relative to the target mobile chassis is generated based on relative state information and compliant control methods. If the target movement mode is an autonomous movement mode, then collect demonstration action data of humans going up and down stairs or up and down a chassis, determine the initial movement strategy based on the demonstration action data and using an imitation learning algorithm, and optimize the initial movement strategy by reinforcement learning based on a preset reward function to obtain the target movement strategy. The execution process of the target action strategy is monitored, and the state changes of the robot and the target mobile chassis are determined based on the monitoring results. The target control parameters are then adjusted based on the state changes until the robot completes the corresponding chassis loading and unloading operations.

[0006] Optionally, determining the target motion pattern of the robot when it is mounted or dismounted from the chassis based on the current environment and the relative state information includes: Determine whether a target assistant exists in the current environment corresponding to the robot; If the target assisting party exists, then the target motion mode of the robot when it is on the upper and lower chassis is characterized as the assisting motion mode; If the target assisting party does not exist, then the target motion mode of the robot when it is on the upper and lower chassis is characterized as an autonomous motion mode.

[0007] Optionally, if the target motion mode is an assisted motion mode, then a target motion strategy for the robot relative to the target mobile chassis is generated based on relative state information and a compliant control method, including: If the target motion mode is an auxiliary motion mode, then the motion trajectory of the robot relative to the target mobile chassis is determined based on the relative state information; The robot's compliance control is activated, and the corresponding target compliance control parameters are determined based on the height difference; The target motion strategy of the robot relative to the target moving chassis is constructed based on the target compliant control parameters and the motion trajectory.

[0008] Optionally, determining the corresponding target compliance control parameters based on the height difference includes: If the height difference is greater than the target threshold, the compliance control parameter corresponding to the compliance control is increased to obtain the target compliance control parameter; If the height difference is not greater than the target threshold, the compliance control parameter is reduced to obtain the target compliance control parameter.

[0009] Optionally, if the target movement mode is an autonomous movement mode, then demonstration action data of humans going up and down stairs or up and down a chassis are collected. Based on the demonstration action data and using an imitation learning algorithm, an initial movement strategy is determined, and the initial movement strategy is optimized by reinforcement learning based on a preset reward function to obtain the target movement strategy, including: If the target motion mode is autonomous motion mode, then determine whether the height of the target mobile chassis is fixed; If the height of the target mobile chassis is fixed, the robot's calibration position in front of the target mobile chassis is determined by using a preset movement command, and the target action strategy that the robot performs relative to the target mobile chassis is determined based on the calibration position. If the height of the target mobile chassis is not fixed, then collect demonstration action data of humans going up and down the steps or up and down the chassis platform, and determine the initial action strategy based on the demonstration action data and using the imitation learning algorithm. The initial action strategy is reinforced by external perturbation factors and a preset reward function to obtain a learned action strategy. The learned action strategy is then mapped into the robot's execution space, and the target action strategy of the robot relative to the target mobile chassis is determined by combining the relative state information.

[0010] Optionally, before monitoring the execution process of the target action strategy, the following steps are included: The identification information affixed to the robot's hips or torso is scanned, and the target position and target orientation of the robot are determined using the scan results. The position and orientation of the target mobile chassis are adjusted based on the target position and the target orientation to obtain an adjusted chassis, and the robot is driven to execute a target motion strategy relative to the adjusted chassis.

[0011] Optional, also includes: If the robot boards the target mobile chassis, the support structure of the adjusted chassis is obtained, and the connection operation between the robot and the adjusted chassis is completed based on the support structure and the target posture of the robot; the target posture is any one of squatting, sitting, or standing.

[0012] Secondly, this application provides a robot motion control device, comprising: The status information acquisition module is used to acquire the relative status information between the robot and the target mobile chassis; the relative status information includes the distance, relative angle, and height difference between the robot and the target mobile chassis; A motion mode determination module is used to determine the target motion mode of the robot when it is on the upper and lower chassis based on the current environment and the relative state information of the robot; the target motion mode includes an assisted motion mode and an autonomous motion mode. The motion strategy generation module is used to generate a target motion strategy for the robot relative to the target mobile chassis based on relative state information and a compliant control method if the target motion mode is an auxiliary motion mode. The action strategy optimization module is used to collect demonstration action data of humans going up and down stairs or up and down a chassis if the target movement mode is an autonomous movement mode, determine an initial action strategy based on the demonstration action data and using an imitation learning algorithm, and optimize the initial action strategy by reinforcement learning based on a preset reward function to obtain a target action strategy. The control parameter adjustment module is used to monitor the execution process of the target action strategy, determine the state changes of the robot and the target mobile chassis based on the monitoring results, and adjust the target control parameters using the state changes until the robot completes the corresponding chassis loading and unloading operations.

[0013] Thirdly, this application provides an electronic device, comprising: Memory, used to store computer programs; A processor is used to execute the computer program to implement the aforementioned robot motion control method.

[0014] Fourthly, this application provides a computer-readable storage medium for storing a computer program, wherein the computer program, when executed by a processor, implements the aforementioned robot motion control method.

[0015] This application acquires relative state information between a robot and a target mobile chassis; the relative state information includes the distance, relative angle, and height difference between the robot and the target mobile chassis; based on the current environment corresponding to the robot and the relative state information, it determines the target motion mode of the robot when moving up and down the chassis; the target motion mode includes an assisted motion mode and an autonomous motion mode; if the target motion mode is an assisted motion mode, it generates a target action strategy for the robot relative to the target mobile chassis based on the relative state information and a compliant control method; if the target motion mode is an autonomous motion mode, it collects demonstration action data of humans moving up and down stairs or up and down the chassis, determines an initial action strategy based on the demonstration action data and using an imitation learning algorithm, and optimizes the initial action strategy by reinforcement learning based on a preset reward function to obtain a target action strategy; it monitors the execution process of the target action strategy, determines the state changes of the robot and the target mobile chassis based on the obtained monitoring results, and adjusts the target control parameters using the state changes until the robot completes the corresponding up and down chassis operation.

[0016] As can be seen from the above, this application acquires the relative state information of the robot and the target mobile chassis, and then determines whether to use an assisted mode or an autonomous mode based on the current environment. Different control strategies are determined based on different motion modes. In assisted mode, a compliant control method is used to determine the robot's target action strategy; in autonomous mode, standard human motion data is collected to learn and imitate the robot, thereby ensuring the naturalness and accuracy of motion execution. This ensures that the robot will not fall due to motion deviation and guarantees a good user experience during human-robot collaboration. During motion execution, state changes are monitored in real time, and core parameters are adjusted accordingly. In this way, even in complex outdoor environments, the robot can accurately move up and down the chassis. By combining imitation learning and reinforcement learning, the robot's robustness in complex environments is improved, eliminating the need to develop separate control logic for different scenarios and reducing R&D and application costs. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention 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 only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0018] Figure 1 This is a flowchart of a robot motion control method disclosed in this application; Figure 2 This is a schematic diagram of the structure of a robot motion control device disclosed in this application; Figure 3 This is a structural diagram of an electronic device disclosed in this application. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention 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, and 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] Currently, existing solutions lack a unified system architecture and multi-mode switching mechanism, and do not fully consider the coordinated adjustment between the chassis and the robot. Furthermore, existing solutions typically rely on manual remote control or fixed motion scripts, making it difficult to adapt to mobile chassis at different heights or positions. These limitations restrict the expansion of practical application scenarios. To address this, this application provides a robot motion control method that can accurately control the robot's movement up and down the chassis, even in complex outdoor environments. It improves the robot's robustness in complex environments by combining imitation learning and reinforcement learning, eliminating the need to develop separate control logic for different scenarios and reducing R&D and application costs.

[0021] See Figure 1 As shown, an embodiment of the present invention discloses a robot motion control method, including: Step S11: Obtain the relative state information of the robot and the target mobile chassis; the relative state information includes the distance, relative angle and height difference between the robot and the target mobile chassis.

[0022] In this embodiment, the distance between the robot and the target mobile chassis is measured, which can be a straight-line distance; and the relative angle and height difference between the robot and the target mobile chassis are measured.

[0023] Step S12: Determine the target motion mode of the robot when it is on the upper and lower chassis based on the current environment and the relative state information corresponding to the robot; the target motion mode includes assisted motion mode and autonomous motion mode.

[0024] In this embodiment, the current environment in which the robot is located is assessed, and the target motion mode of the robot when moving up and down the chassis is determined based on the assessment result. If a target assisting party exists in the current environment, the target motion mode is an assisted motion mode; if the target assisting party does not exist in the current environment, the target motion mode is an autonomous motion mode; the target assisting party is a person or object used to assist the robot in moving up and down the chassis.

[0025] Specifically, determining the target motion mode of the robot when it is moving up and down the chassis based on the current environment corresponding to the robot and the relative state information includes: determining whether there is a target auxiliary party in the current environment corresponding to the robot; if the target auxiliary party exists, the target motion mode of the robot when it is moving up and down the chassis is characterized as an assisted motion mode; if the target auxiliary party does not exist, the target motion mode of the robot when it is moving up and down the chassis is characterized as an autonomous motion mode.

[0026] Step S13: If the target motion mode is an auxiliary motion mode, then generate a target motion strategy for the robot relative to the target moving chassis based on relative state information and compliant control method.

[0027] In this embodiment, if it is an assisted motion mode, the vertical movement trajectory of the robot on and off the target moving chassis is determined based on the relative state information, such as which leg to step on first and how high to lift. The robot's impedance control or compliance control is activated to ensure compliance when the robot contacts the target assisting party, and to allow the target assisting party to apply guiding force to the robot. Then, the target compliance control parameters are adjusted based on the height difference; the target compliance control parameters include impedance control parameters.

[0028] Specifically, if the target motion mode is an auxiliary motion mode, then generating a target motion strategy for the robot relative to the target mobile chassis based on relative state information and a compliant control method includes: if the target motion mode is an auxiliary motion mode, determining the motion trajectory of the robot relative to the target mobile chassis based on the relative state information; enabling compliant control of the robot and determining corresponding target compliant control parameters based on the height difference; and constructing the target motion strategy for the robot relative to the target mobile chassis based on the target compliant control parameters and the motion trajectory.

[0029] It is understood that when the height difference is greater than the target threshold, it indicates that the target mobile chassis is relatively high. Therefore, the compliance control parameter is increased to avoid the robot generating excessive reaction force on the target auxiliary party, ensuring both that the robot does not fall over and that the target auxiliary party operates safely and effortlessly. Conversely, the compliance control parameter is decreased. The target threshold can be set according to actual conditions. Specifically, determining the corresponding target compliance control parameter based on the height difference includes: if the height difference is greater than the target threshold, increasing the compliance control parameter to obtain the target compliance control parameter; if the height difference is not greater than the target threshold, decreasing the compliance control parameter to obtain the target compliance control parameter.

[0030] Step S14: If the target movement mode is an autonomous movement mode, collect demonstration action data of humans going up and down stairs or up and down a chassis, determine the initial movement strategy based on the demonstration action data and using an imitation learning algorithm, and optimize the initial movement strategy by reinforcement learning based on a preset reward function to obtain the target movement strategy.

[0031] In this embodiment, if the target motion mode is an autonomous motion mode, it is further determined whether the target mobile chassis has a definite height or a fixed height. If so, the robot's calibration position in front of the target mobile chassis is determined by remote control or high-level command. The robot is pre-trained to perform up-and-down movement at a fixed height / preset height, and then the robot's calibration position in front of the target mobile chassis is determined using a preset movement command. If the robot is climbing onto the target mobile chassis, the robot is moved to the calibration position and the target motion strategy is executed. If the robot is dismounting from the target mobile chassis, the robot is moved to the calibration position based on the target motion strategy. If not, then demonstration action data of humans going up and down stairs or onto and off a chassis platform are collected, and the robot is trained to learn human movement rhythms and postures using an imitation learning algorithm, which can be the BeyondMimic training framework. Various unexpected situations are introduced, including slight slippage on the ground and slight chassis misalignment. The robot is iteratively trained using reinforcement learning, with corresponding preset reward functions set during training, until the target conditions are met. The trained target action policy is then mapped into the robot's execution space, resulting in a more natural target action policy. Random perturbations can be applied to the robot or the target moving chassis during training to improve the robustness of the action policy.

[0032] In one specific implementation, the preset reward function can be a reward function determined based on the robot's landing position. Specifically, a target landing position is set for the robot, and it is determined whether the difference between the actual landing position corresponding to the reinforcement learning and the target landing position is greater than a target difference threshold. If it is greater than the target difference threshold, a negative reward is given; if it is not greater than the target difference threshold, a positive reward is given. Through this reward mechanism, the robot can accurately land its feet in a predetermined area during the process of getting on and off the chassis, thereby ensuring the stability of the robot's center of gravity during the process and improving the success rate of the action. Similarly, the preset reward function can be a reward function determined based on action imitation error, a reward function determined based on posture stability, or a reward function based on energy consumption.

[0033] Specifically, if the target motion mode is an assisted motion mode, a target action strategy for the robot relative to the target mobile chassis is generated based on relative state information and a compliant control method. This includes: if the target motion mode is an autonomous motion mode, determining whether the height of the target mobile chassis is fixed; if the height of the target mobile chassis is fixed, determining the robot's calibration position in front of the target mobile chassis using a preset movement command, and determining the target action strategy for the robot to execute relative to the target mobile chassis based on the calibration position; if the height of the target mobile chassis is not fixed, collecting demonstration action data of humans going up and down stairs or up and down chassis platforms, determining an initial action strategy based on the demonstration action data and using an imitation learning algorithm; performing reinforcement learning on the initial action strategy using external perturbation factors and a preset reward function to obtain a learned action strategy, mapping the learned action strategy to the robot's execution space, and determining the robot's target action strategy relative to the target mobile chassis in conjunction with the relative state information.

[0034] In addition, when the height of the target mobile chassis is not fixed, the robot's environmental perception module can identify the position, height and relative posture of the target mobile chassis, and adjust the robot's gait parameters and motion strategy based on the identification results, and determine the target motion strategy based on the adjusted gait parameters and motion strategy; the environmental perception module is a module based on LiDAR equipment or RGB-D camera.

[0035] Step S15: Monitor the execution process of the target action strategy, determine the state changes of the robot and the target mobile chassis based on the monitoring results, and adjust the target control parameters using the state changes until the robot completes the corresponding chassis loading and unloading operations.

[0036] It is understood that, regardless of the target motion mode, before executing the target motion strategy, a collaborative alignment operation between the robot and the target mobile chassis is performed. This involves scanning the markings affixed to the robot's hip or torso, and adjusting the position and orientation of the target mobile chassis based on the scan results to obtain an adjusted chassis. The markings can be QR codes, AprilTags (i.e., visual reference systems), and visual markers. By fine-tuning the positional accuracy of the target mobile chassis, posture deviations when the robot moves up and down the chassis can be reduced. Specifically, before monitoring the execution of the target motion strategy, the process includes: scanning the markings affixed to the robot's hip or torso, and using the scan results to determine the robot's target position and target orientation; adjusting the position and orientation of the target mobile chassis based on the target position and target orientation to obtain an adjusted chassis, and then driving the robot to execute the target motion strategy relative to the adjusted chassis.

[0037] In this embodiment, the execution process of the target motion strategy is monitored in real time, that is, the robot's posture, forces, and the relative state between the robot and the target mobile chassis are monitored. If any deviation is detected, such as the robot tilting, increased guidance force from the target auxiliary side, or slight chassis movement, the target control parameters are adjusted. These target control parameters include the robot's walking speed and compliance control parameters to ensure the stability and safety of the robot's movement. Furthermore, after the robot successfully mounts the target mobile chassis, the target mobile chassis automatically extends a support rod or seat structure to provide additional support to the robot, allowing the robot to choose one or more postures, such as squatting, sitting, or standing, based on the support structure provided by the target mobile chassis. Specifically, if the robot mounts the target mobile chassis, the adjusted chassis support structure is obtained, and the connection operation between the robot and the adjusted chassis is completed based on the support structure and the robot's target posture; the target posture is any one of squatting, sitting, or standing.

[0038] As can be seen from the above, this application acquires the relative state information of the robot and the target mobile chassis, and then determines whether to use an assisted mode or an autonomous mode based on the current environment. Different control strategies are determined based on different motion modes. In assisted mode, a compliant control method is used to determine the robot's target action strategy; in autonomous mode, standard human motion data is collected to learn and imitate the robot, thereby ensuring the naturalness and accuracy of motion execution. This ensures that the robot will not fall due to motion deviation and guarantees a good user experience during human-robot collaboration. During motion execution, state changes are monitored in real time, and core parameters are adjusted accordingly. In this way, even in complex outdoor environments, the robot can accurately move up and down the chassis. By combining imitation learning and reinforcement learning, the robot's robustness in complex environments is improved, eliminating the need to develop separate control logic for different scenarios and reducing R&D and application costs.

[0039] Accordingly, see Figure 2 As shown, this application also provides a robot motion control device, including: The status information acquisition module 11 is used to acquire the relative status information between the robot and the target mobile chassis; the relative status information includes the distance, relative angle and height difference between the robot and the target mobile chassis; The motion mode determination module 12 is used to determine the target motion mode of the robot when it is on the upper and lower chassis based on the current environment and the relative state information of the robot; the target motion mode includes an assisted motion mode and an autonomous motion mode. The motion strategy generation module 13 is used to generate a target motion strategy for the robot relative to the target mobile chassis based on relative state information and a compliant control method if the target motion mode is an auxiliary motion mode. The action strategy optimization module 14 is used to collect demonstration action data of humans going up and down stairs or up and down chassis if the target movement mode is an autonomous movement mode, determine the initial action strategy based on the demonstration action data and using an imitation learning algorithm, and optimize the initial action strategy by reinforcement learning based on a preset reward function to obtain the target action strategy. The control parameter adjustment module 15 is used to monitor the execution process of the target action strategy, determine the state changes of the robot and the target mobile chassis based on the monitoring results, and adjust the target control parameters in the control strategy using the state changes until the robot completes the corresponding chassis loading and unloading operations.

[0040] In some specific embodiments, the motion pattern determination module 12 may specifically include: An environment determination unit is used to determine whether a target assistant exists in the current environment corresponding to the robot. The first mode determination unit is used to characterize the target motion mode of the robot when it is on the upper and lower chassis as an auxiliary motion mode if the target auxiliary party exists. The second mode determination unit is used to characterize the target motion mode of the robot when it is on the upper and lower chassis as an autonomous motion mode if the target auxiliary party does not exist.

[0041] In some specific embodiments, the action strategy generation module 13 may specifically include: A motion trajectory determination unit is used to determine the motion trajectory of the robot relative to the target mobile chassis based on the relative state information if the target motion mode is an auxiliary motion mode. The control parameter determination submodule is used to enable the compliant control of the robot and determine the corresponding target compliant control parameters based on the height difference. The target strategy determination unit is used to construct the target motion strategy of the robot relative to the target moving chassis based on the target compliant control parameters and the motion trajectory.

[0042] In some specific implementations, the control parameter determination submodule may specifically include: The control parameter increasing unit is used to increase the compliance control parameter corresponding to the compliance control if the height difference is greater than the target threshold, so as to obtain the target compliance control parameter. The control parameter reduction unit is used to reduce the compliance control parameter if the height difference is not greater than the target threshold, so as to obtain the target compliance control parameter.

[0043] In some specific embodiments, the action strategy optimization module 14 may specifically include: The height determination unit is used to determine whether the height of the target mobile chassis is fixed if the target's motion mode is an autonomous motion mode. The target strategy execution unit is used to determine the calibration position of the robot in front of the target mobile chassis using a preset movement command if the height of the target mobile chassis is fixed, and to determine the target action strategy that the robot will execute relative to the target mobile chassis based on the calibration position. The initial strategy determination unit is used to collect demonstration action data of humans going up and down stairs or up and down the platform if the height of the target mobile chassis is not fixed, and to determine the initial action strategy based on the demonstration action data and using an imitation learning algorithm. The action policy mapping unit is used to perform reinforcement learning on the initial action policy using external perturbation factors and a preset reward function to obtain the learned action policy, map the learned action policy to the robot's execution space, and combine the relative state information to determine the target action policy of the robot relative to the target mobile chassis.

[0044] In some specific embodiments, the robot motion control device may further include: The identification information scanning unit is used to scan the identification information pasted on the robot's crotch or torso, and use the obtained scanning results to determine the robot's target position and target orientation; The position adjustment unit is used to adjust the position and orientation of the target mobile chassis based on the target position and the target orientation to obtain the adjusted chassis, and drive the robot to execute the target motion strategy relative to the adjusted chassis.

[0045] In some specific embodiments, the robot motion control device may further include: The support result acquisition unit is used to acquire the support structure of the adjusted chassis if the robot boards the target mobile chassis, and to complete the connection operation between the robot and the adjusted chassis based on the support structure and the target posture of the robot; the target posture is any one of squatting, sitting, or standing.

[0046] Furthermore, embodiments of this application also disclose an electronic device, Figure 3 This is a structural diagram of an electronic device 20 according to an exemplary embodiment. The content of the diagram should not be construed as limiting the scope of this application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 stores a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the robot motion control method disclosed in any of the foregoing embodiments. Furthermore, the electronic device 20 in this embodiment may specifically be an electronic computer.

[0047] In this embodiment, the power supply 23 is used to provide operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and external devices, and the communication protocol it follows can be any communication protocol applicable to the technical solution of this application, and is not specifically limited here; the input / output interface 25 is used to acquire external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs, and is not specifically limited here.

[0048] In addition, the memory 22, as a carrier for resource storage, can be a read-only memory, random access memory, disk or optical disk, etc. The resources stored thereon can include operating system 221, computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0049] The operating system 221 is used to manage and control the various hardware devices on the electronic device 20 and the computer program 222, which may be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program capable of performing the robot motion control method executed by the electronic device 20 as disclosed in any of the foregoing embodiments, the computer program 222 may further include a computer program capable of performing other specific tasks.

[0050] Furthermore, this application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, it implements the aforementioned disclosed robot motion control method. Specific steps of this method can be found in the corresponding content disclosed in the foregoing embodiments, and will not be repeated here.

[0051] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

[0052] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0053] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0054] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0055] The technical solutions provided in this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A robot motion control method, characterized in that, include: Acquire the relative state information between the robot and the target mobile chassis; the relative state information includes the distance, relative angle, and height difference between the robot and the target mobile chassis; The target motion mode of the robot when it is on the upper and lower chassis is determined based on the current environment and the relative state information of the robot; the target motion mode includes assisted motion mode and autonomous motion mode. If the target motion mode is an auxiliary motion mode, then a target motion strategy for the robot relative to the target mobile chassis is generated based on relative state information and compliant control methods. If the target movement mode is an autonomous movement mode, then collect demonstration action data of humans going up and down stairs or up and down a chassis, determine the initial movement strategy based on the demonstration action data and using an imitation learning algorithm, and optimize the initial movement strategy by reinforcement learning based on a preset reward function to obtain the target movement strategy. The execution process of the target action strategy is monitored, and the state changes of the robot and the target mobile chassis are determined based on the monitoring results. The target control parameters are then adjusted based on the state changes until the robot completes the corresponding chassis loading and unloading operations.

2. The robot motion control method according to claim 1, characterized in that, Determining the target motion pattern of the robot when it is mounted or dismounted from the chassis based on the current environment and the relative state information includes: Determine whether a target assistant exists in the current environment corresponding to the robot; If the target assisting party exists, then the target motion mode of the robot when it is on the upper and lower chassis is characterized as the assisting motion mode; If the target assisting party does not exist, then the target motion mode of the robot when it is on the upper and lower chassis is characterized as an autonomous motion mode.

3. The robot motion control method according to claim 1, characterized in that, If the target motion mode is an assisted motion mode, then a target motion strategy for the robot relative to the target mobile chassis is generated based on relative state information and a compliant control method, including: If the target motion mode is an auxiliary motion mode, then the motion trajectory of the robot relative to the target mobile chassis is determined based on the relative state information; The robot's compliance control is activated, and the corresponding target compliance control parameters are determined based on the height difference; The target motion strategy of the robot relative to the target moving chassis is constructed based on the target compliant control parameters and the motion trajectory.

4. The robot motion control method according to claim 3, characterized in that, The determination of the corresponding target compliance control parameters based on the height difference includes: If the height difference is greater than the target threshold, the compliance control parameter corresponding to the compliance control is increased to obtain the target compliance control parameter; If the height difference is not greater than the target threshold, the compliance control parameter is reduced to obtain the target compliance control parameter.

5. The robot motion control method according to claim 1, characterized in that, If the target movement mode is an autonomous movement mode, then demonstration action data of humans going up and down stairs or onto and off a platform are collected. Based on the demonstration action data and using an imitation learning algorithm, an initial movement strategy is determined. Then, based on a preset reward function, the initial movement strategy is optimized through reinforcement learning to obtain the target movement strategy, including: If the target motion mode is autonomous motion mode, then determine whether the height of the target mobile chassis is fixed; If the height of the target mobile chassis is fixed, the robot's calibration position in front of the target mobile chassis is determined by using a preset movement command, and the target action strategy that the robot performs relative to the target mobile chassis is determined based on the calibration position. If the height of the target mobile chassis is not fixed, then collect demonstration action data of humans going up and down the steps or up and down the chassis platform, and determine the initial action strategy based on the demonstration action data and using the imitation learning algorithm. The initial action strategy is reinforced by external perturbation factors and a preset reward function to obtain a learned action strategy. The learned action strategy is then mapped into the robot's execution space, and the target action strategy of the robot relative to the target mobile chassis is determined by combining the relative state information.

6. The robot motion control method according to any one of claims 1 to 5, characterized in that, Before monitoring the execution process of the target action strategy, the following steps are included: The identification information affixed to the robot's hips or torso is scanned, and the target position and target orientation of the robot are determined using the scan results. The position and orientation of the target mobile chassis are adjusted based on the target position and the target orientation to obtain an adjusted chassis, and the robot is driven to execute a target motion strategy relative to the adjusted chassis.

7. The robot motion control method according to claim 6, characterized in that, Also includes: If the robot boards the target mobile chassis, the support structure of the adjusted chassis is obtained, and the connection operation between the robot and the adjusted chassis is completed based on the support structure and the target posture of the robot; the target posture is any one of squatting, sitting, or standing.

8. A robot motion control device, characterized in that, include: The status information acquisition module is used to acquire the relative status information between the robot and the target mobile chassis; the relative status information includes the distance, relative angle, and height difference between the robot and the target mobile chassis; A motion mode determination module is used to determine the target motion mode of the robot when it is on the upper and lower chassis based on the current environment and the relative state information of the robot; the target motion mode includes an assisted motion mode and an autonomous motion mode. The motion strategy generation module is used to generate a target motion strategy for the robot relative to the target mobile chassis based on relative state information and a compliant control method if the target motion mode is an auxiliary motion mode. The action strategy optimization module is used to collect demonstration action data of humans going up and down stairs or up and down a chassis if the target movement mode is an autonomous movement mode, determine an initial action strategy based on the demonstration action data and using an imitation learning algorithm, and optimize the initial action strategy by reinforcement learning based on a preset reward function to obtain a target action strategy. The control parameter adjustment module is used to monitor the execution process of the target action strategy, determine the state changes of the robot and the target mobile chassis based on the monitoring results, and adjust the target control parameters using the state changes until the robot completes the corresponding chassis loading and unloading operations.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program to implement the robot motion control method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, Used to store a computer program, wherein the computer program, when executed by a processor, implements the robot motion control method as described in any one of claims 1 to 7.

Images