Robot control method, robot control system, and robot

Abstract

Disclosed are a robot motion control method, a robot control system and a robot. The robot comprises a manipulator, and the method comprises: controlling the manipulator to collide with a target object; determining a post-collision motion type of the target object, the post-collision motion type of the target object comprising: a disengagement motion type and a non-disengagement motion type; and controlling the manipulator to grasp the target object based on the post-collision motion type of the target object.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of artificial intelligence and robots, and more particularly to a robot control method, a robot control system and a robot. BACKGROUND

[0002] With the wide application of artificial intelligence and robot technology in civil and commercial fields, robots based on artificial intelligence and robot technology play an increasingly important role in intelligent industrial control, intelligent home control and other fields, and also face higher requirements.

[0003] Currently, when a robot grasps a target object, a static grasping strategy is usually adopted, that is, in the case of robot staticity or quasi-staticity (very low robot speed), the robot uses its dexterous end tool (such as a manipulator) to grasp the object. In this case, on the one hand, the robot performing a motion task (in which the robot has a corresponding motion speed and acceleration) will first stop or slow down for grasping, and then gradually accelerate to the required speed after grasping the object, and the static grasping process affects the execution of the original motion task of the robot; on the other hand, the grasping action is not continuous and the fluency of the grasping operation is low in the static grasping process.

[0004] Therefore, it is necessary to implement grasping of a target object by a robot in a motion process via a dynamic grasping strategy under the premise of good grasping of the target object by the robot, so as to reduce the influence of the grasping operation on the motion task of the robot itself and realize a smooth and continuous target object grasping process. SUMMARY

[0005] To solve the above problems, the present application provides a robot control method, a robot control system and a robot. The robot control method, the robot control system and the robot provided by the present application can implement grasping of a target object by a robot in a motion process via a dynamic grasping strategy under the premise of good grasping of the target object by the robot, so as to reduce the influence of the grasping operation on the motion task of the robot itself and realize a smooth and continuous target object grasping process.

[0006] According to an aspect of the present application, a robot control method is provided, the robot comprising a manipulator, and the method comprising: controlling the manipulator to collide with a target object; determining a post-collision motion type of the target object, the post-collision motion type of the target object comprising: a disengagement motion type and a non-disengagement motion type; and controlling the manipulator to grasp the target object based on the post-collision motion type of the target object.

[0007] In some embodiments, the controlling the robot arm to collide with the target object comprises: determining motion state information of the target object at a beginning time of the collision; obtaining expected motion state information of the target object at an ending time of the collision, and expected motion state information of the robot arm at the ending time of the collision; determining target motion state information of the robot arm at the beginning time of the collision based on a collision motion model, based on the motion state information of the target object at the beginning time of the collision, the expected motion state information of the target object at the ending time of the collision, and the expected motion state information of the robot arm at the ending time of the collision; and controlling the robot arm to collide with the target object based on the target motion state information of the robot arm at the beginning time of the collision.

[0008] In some embodiments, the motion state information of the target object at the beginning time of the collision comprises: a center-of-mass motion position, a center-of-mass motion velocity, an attitude angle, and an attitude angular velocity of the target object at the beginning time of the collision; the expected motion state information of the target object at the ending time of the collision comprises: an expected velocity of a collision point of the target object at the ending time of the collision, and an expected center-of-mass motion position, an expected center-of-mass motion velocity, an expected attitude angle, and an expected attitude angular velocity of the target object at the ending time of the collision; the expected motion state information of the robot arm at the ending time of the collision comprises: an expected center-of-mass motion position, an expected center-of-mass motion velocity, an expected attitude angle, and an expected attitude angular velocity of the robot arm at the ending time of the collision; and the target motion state information of the robot arm at the beginning time of the collision comprises: a target position, a target velocity, a target attitude angle, and a target attitude angular velocity of the robot arm at the beginning time of the collision.

[0009] In some embodiments, the determining target motion state information of the robot arm at the beginning time of the collision based on a collision motion model, based on the motion state information of the target object at the beginning time of the collision, the expected motion state information of the target object at the ending time of the collision, and the expected motion state information of the robot arm at the ending time of the collision comprises: determining an expected collision impulse of the target object based on the motion state information of the target object at the beginning time of the collision and the expected motion state information of the target object at the ending time of the collision; determining an expected motion velocity of a collision point of the target object at the beginning time of the collision based on the expected collision impulse of the target object, the expected motion state information of the target object at the ending time of the collision, and the expected motion state information of the robot arm at the ending time of the collision; and determining the target motion state information of the robot arm at the beginning time of the collision based on the expected motion velocity of the collision point of the target object at the beginning time of the collision and the motion state information of the target object at the beginning time of the collision.

[0010] In some embodiments, determining the post-collision motion type of the target object comprises: obtaining motion state information of the target object at a time point when the collision ends and motion state information of the robot at the time point when the collision ends; comparing the motion state information of the target object at the time point when the collision ends with the motion state information of the robot at the time point when the collision ends; and determining the post-collision motion type of the target object based on the comparison result.

[0011] In some embodiments, controlling the robot to grasp the target object based on the post-collision motion type of the target object comprises: based on the post-collision motion type of the target object, controlling the robot to approach the target object so that the target object is on the robot and the robot is in a same pose as the target object; and controlling the robot to perform a gripping operation to grasp the target object.

[0012] In some embodiments, when the target post-collision motion type is the non-detached motion type, controlling the robot to approach the target object based on the post-collision motion type of the target object comprises: obtaining post-collision motion state information of the target object; obtaining expected post-collision motion state information of the target object corresponding to the non-detached motion type; based on the post-collision motion state information of the target object and the expected post-collision motion state information of the target object corresponding to the non-detached motion type, determining expected approach information of the robot acting on the target object according to a non-detached motion model, the expected approach information comprising: expected contact force, expected contact position, and expected contact pose; obtaining current approach information of the robot acting on the target object, and controlling the robot to approach the target object based on the current approach information and the expected approach information.

[0013] In some embodiments, when the target post-collision motion type is the detached motion type, controlling the robot to approach the target object based on the post-collision motion type of the target object comprises: obtaining post-collision motion state information of the target object, and based on the post-collision motion state information of the target object, generating a post-collision motion trajectory and pose estimation of the target object according to a detached motion model; based on the motion trajectory and pose estimation of the target object, determining expected approach motion state information of the robot, the expected approach motion state information comprising: expected approach position information, expected approach velocity information, and expected approach pose information; and controlling the robot to approach the target object based on the expected approach motion state information of the robot.

[0014] In some embodiments, the robot includes a plurality of fingers, and the robot is controlled to perform a grasping operation to grasp the target object includes: controlling the plurality of fingers to perform a grasping motion to approach each other; obtaining contact force data of an end of at least one of the plurality of fingers; and stopping the grasping motion when the contact force data is greater than or equal to a preset contact force threshold.

[0015] In some embodiments, the current contact information includes: current contact force of the robot acting on the target object, and obtaining the current contact force of the robot acting on the target object includes: obtaining current motion state information of the target object; determining external force acting on the target object based on the current motion state information of the target object based on a dynamics equation; determining the current contact force based on the external force.

[0016] According to another aspect of the present disclosure, a robot motion control system is provided, the robot includes a robot arm, and the system includes: a collision control module configured to control the robot arm to collide with a target object; a post-collision motion type determination module configured to determine a post-collision motion type of the target object, the post-collision motion type of the target object including: a disengagement motion type and a non-disengagement motion type; a grasping module configured to control the robot arm to grasp the target object based on the post-collision motion type of the target object.

[0017] In some embodiments, the collision control module includes: an object motion state determination module configured to determine motion state information of the target object at a collision initial time; an expected motion state obtaining module configured to obtain expected motion state information of the target object at a collision end time, expected motion state information of the robot arm at the collision end time; a robot arm collision information generation module configured to determine target motion state information of the robot arm at the collision initial time based on the motion state information of the target object at the collision initial time, the expected motion state information of the target object at the collision end time, the expected motion state information of the robot arm at the collision end time based on a collision motion model; a collision generation module configured to control the robot arm to collide with the target object based on the target motion state information of the robot arm at the collision initial time.

[0018] In some embodiments, the motion state information of the target object at the initial time of collision includes: a mass center motion position, a mass center motion velocity, an attitude angle, and an attitude angular velocity of the target object at the initial time of collision; the desired motion state information of the target object at the end time of collision includes: a desired velocity of the collision point of the target object at the end time of collision, and a desired mass center motion position, a desired mass center motion velocity, a desired attitude angle, and a desired attitude angular velocity of the target object at the end time of collision; the desired motion state information of the robot at the end time of collision includes: a desired mass center motion position, a desired mass center motion velocity, a desired attitude angle, and a desired attitude angular velocity of the robot at the end time of collision; and the target motion state information of the robot at the initial time of collision includes: a target position, a target velocity, a target attitude angle, and a target attitude angular velocity of the robot at the initial time of collision.

[0019] In some embodiments, the robot collision information generation module includes: a collision impulse calculation module configured to determine a desired collision impulse of the target object based on the motion state information of the target object at the initial time of collision and the desired motion state information of the target object at the end time of collision; a collision point initial velocity determination module configured to determine a desired motion velocity of the collision point of the target object at the initial time of collision based on the desired collision impulse of the target object, the desired motion state information of the target object at the end time of collision, and the desired motion state information of the robot at the end time of collision; and a robot initial velocity determination module configured to determine the target motion state information of the robot at the initial time of collision based on the desired motion velocity of the collision point of the target object at the initial time of collision and the motion state information of the target object at the initial time of collision.

[0020] In some embodiments, the post-collision motion type determination module includes: a collision end motion state acquisition module configured to acquire the motion state information of the collision point of the target object at the end time of collision and the motion state information of the robot at the end time of collision; a motion state comparison module configured to compare the motion state information of the collision point of the target object at the end time of collision with the motion state information of the robot at the end time of collision; and a motion type determination module configured to determine the post-collision motion type of the target object based on a comparison result.

[0021] In some embodiments, the grabbing module includes: a picking operation module configured to control the robot to pick the target object based on the post-collision motion type of the target object, so that the target object is located on the robot and the robot is consistent with the attitude of the target object; and a clenching operation module configured to control the robot to perform a clenching operation to grab the target object.

[0022] According to another aspect of the present disclosure, a robot is provided, the robot comprising a robot hand, and wherein the robot comprises a robot motion control system as described above, and the control of the robot is achieved by a robot motion control method as described above. BRIEF DESCRIPTION OF DRAWINGS

[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present disclosure, and other drawings can also be obtained by those skilled in the art without any creative work on the basis of these drawings. The following drawings are not necessarily drawn in proportion to the actual size, and the focus is on showing the main idea of the present disclosure.

[0024] Figure 1 An exemplary flowchart of a robot control method 100 according to an embodiment of the present disclosure is shown;

[0025] Figure 2 An exemplary flowchart of a process S101 of controlling the robot hand to collide with the target object according to an embodiment of the present disclosure is shown;

[0026] Figure 3A An exemplary flowchart of a process S1013 of determining the target motion state information of the robot hand at the initial time of collision based on the collision motion model according to an embodiment of the present disclosure is shown;

[0027] Figure 3B A schematic diagram of the collision process of the robot hand and the target object is shown;

[0028] Figure 4 An exemplary flowchart of a process S103 of controlling the robot hand to grasp the target object according to an embodiment of the present disclosure is shown;

[0029] Figure 5A A schematic diagram of the object in the disengagement motion type according to an embodiment of the present disclosure is shown;

[0030] Figure 5B An exemplary flowchart of a process S1031A of controlling the robot hand to grasp the target object in the case that the target post-collision motion type is the disengagement motion type according to an embodiment of the present disclosure is shown;

[0031] Figure 5C A schematic diagram of the robot hand grasping the target object according to an embodiment of the present disclosure is shown;

[0032] Figure 6A A schematic diagram of the motion process of the target object in the non-disengagement motion type according to an embodiment of the present disclosure is shown;

[0033] Figure 6B An exemplary flow chart illustrating a process S1031B of controlling the robot to pick up the target object after the target object colliding according to an embodiment of the present disclosure is shown in FIG. 10B;

[0034] Figure 7 An exemplary schematic diagram illustrating the robot performing the preset motion according to an embodiment of the present disclosure is shown in FIG. 11A;

[0035] Figure 8 An exemplary schematic diagram illustrating the robot picking up the target object during the process of performing the preset motion according to an embodiment of the present disclosure is shown in FIG. 11B;

[0036] Figure 9 An exemplary block diagram of the self-balancing robot control system 600 according to an embodiment of the present disclosure is shown in FIG. 12. DETAILED DESCRIPTION

[0037] The technical solutions in the embodiments of the present disclosure will be described clearly and completely below with reference to the drawings. Obviously, the described embodiments are only part of the embodiments of the present disclosure, but not all the embodiments. Based on the embodiments of the present disclosure, all other embodiments obtained by those skilled in the art without creative work fall within the scope of the present disclosure.

[0038] As shown in the present application and claims, unless the context clearly indicates otherwise, the words “one”, “an”, “a”, and / or “the” do not mean to specify a single number, but can also include a plurality. Generally, the terms “comprise” and “include” only indicate the inclusion of the steps and elements explicitly identified, and these steps and elements do not constitute an exclusive list, and the method or device can also include other steps or elements.

[0039] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, however, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0040] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed in sequence. Instead, various steps can be processed in reverse order or simultaneously, as needed. Meanwhile, other operations can be added to these processes, or one or more steps of the operations can be removed from these processes.

[0041] Artificial Intelligence (AI) is the theory, method, technology and application system of using digital computer or machine controlled by digital computer to simulate, extend and expand human intelligence, perceive environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technology of computer science, which attempts to understand the essence of intelligence and produce a new intelligent machine that can react in a similar way to human intelligence. Artificial intelligence is to study the design principles and implementation methods of various intelligent machines, so that the machine has the functions of perception, reasoning and decision-making.

[0042] Artificial intelligence technology is a comprehensive discipline, involving a wide range of fields, both hardware and software technologies. Artificial intelligence basic technologies generally include technologies such as sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction system, mechatronics, etc. Artificial intelligence software technology mainly includes computer vision technology, speech processing technology, natural language processing technology, and machine learning / deep learning, etc.

[0043] With the wide application of artificial intelligence and robot technology in civil and commercial fields, robots based on artificial intelligence and robot technology play an increasingly important role in intelligent industrial control, intelligent home control and other fields, and also face higher requirements.

[0044] The current robot usually adopts a static grasping strategy when grasping a target object, that is, in the case of static or quasi-static (very low speed) of the robot, the robot uses its dexterous end tool (such as a manipulator) to grasp the object. In this case, on the one hand, the robot performing a motion task (in which the robot has a corresponding motion speed and acceleration) will first stop or slow down for grasping, and then gradually accelerate to the required speed after grasping the object, the static grasping process affects the execution of the original motion task of the robot; on the other hand, the grasping action is not continuous and the fluency of the grasping operation is low in the static grasping process.

[0045] Based on the above, the present application proposes a robot control method based on artificial intelligence, which controls the collision of the manipulator with the target object first, and then controls the dynamic grasping strategy of the manipulator grasping the target object based on the post-collision motion type of the target object, realizes the grasping of the target object by the robot in the motion process, reduces the influence of the grasping operation on the motion task of the robot itself, and realizes a smooth and continuous target object grasping process.

[0046] Figure 1An exemplary flowchart of a robot control method 100 according to an embodiment of the present disclosure is shown. The robot, for example, includes a manipulator, which refers to a component of the robot used to interact with a target object to achieve a grasping process of the target object.

[0047] According to actual needs, the manipulator may, for example, have a palm base and a plurality of finger-shaped parts extending outward from the palm base, or the manipulator may have other forms and structures. Embodiments of the present disclosure are not limited by the specific structure and shape of the manipulator.

[0048] Referring to Figure 1 First, in step S101, the manipulator is controlled to collide with the target object.

[0049] It should be understood that the target object refers to an object that the robot aims to grasp. The target object may, for example, be in a stationary or moving state (e.g., a uniform motion state, a uniform acceleration motion state, or a variable acceleration motion state), and embodiments of the present disclosure are not limited by the motion state of the target object.

[0050] The process of controlling the manipulator to collide with the target object may, for example, be: determining motion state information of the target object at a collision initial time; obtaining desired motion state information of the target object at a collision end time, desired motion state information of the manipulator at the collision end time; determining target motion state information of the manipulator at the collision initial time based on the motion state information of the target object at the collision initial time, the desired motion state information of the target object at the collision end time, the desired motion state information of the manipulator at the collision end time, based on a collision motion model; and controlling the manipulator to collide with the target object based on the target motion state information of the manipulator at the collision initial time. Alternatively, the manipulator may be controlled to collide with the target object in other ways according to actual needs.

[0051] After controlling the manipulator to collide with the target object, in step S102, the post-collision motion type of the target object is determined, including a disengagement motion type and a non-disengagement motion type.

[0052] The post-collision motion type of the target object refers to the motion type of the target object relative to the manipulator after colliding with the manipulator. The disengagement motion type refers to the target object flying away from the manipulator after colliding with the manipulator. The non-disengagement motion type refers to the target object remaining in contact with the manipulator at at least one contact point after colliding with the manipulator, and having the same motion state as the manipulator at the contact point (i.e., both are in a coordinated motion state at the contact point).

[0053] and wherein the process of determining the post-collision motion type of the target object can be implemented, for example, by first obtaining the motion state information of the collision point of the target object at the end of the collision, and then comparing the motion state information of the collision point of the target object at the end of the collision with the motion state information of the robot at the end of the collision, and determining the post-collision motion type of the target object based on the comparison result. For example, the motion speed of the collision point of the target object at the end of the collision can be compared with the motion speed of the robot, and if the motion speed of the collision point of the target object is greater than the motion speed of the robot, it is determined to be the disengaged motion type; otherwise, it is determined to be the non-disengaged motion type.

[0054] It should be understood that the above only gives an exemplary method of determining the post-collision motion type of the target object, and the post-collision motion type of the target object can also be determined based on other manners according to actual needs. For example, the motion state information of the collision point of the target object at the end of the collision can be compared with the preset motion state information, and the post-collision motion type of the target object can be determined based on the comparison result.

[0055] After determining the post-collision motion type of the target object, in step S103, the robot is controlled to grasp the target object based on the post-collision motion type of the target object.

[0056] For example, the robot can first be controlled to perform a picking step on the target object based on the post-collision motion type of the target object, so that the target object is in a picked state; and then the robot is controlled to perform a gripping operation to grasp the target object.

[0057] Based on the above, in the present application, by setting a dynamic grasping strategy, specifically, by first controlling the robot to collide with the target object, determining the post-collision motion type of the target object, and controlling the robot to grasp the target object based on the determined post-collision motion type of the target object, the robot achieves grasping of the target object during motion. Compared with the current target object grasping process based on a static grasping strategy, on the one hand, by adopting a dynamic grasping strategy based on collision, the present application reduces the influence of the grasping operation on the motion task of the robot, so that the robot can achieve good and reliable grasping of the target object without significantly reducing the speed or stopping the motion, so that the process of grasping the target object does not affect the motion task of the robot, which is conducive to the robot to balance the grasping of the target object and the execution of its own motion task. On the other hand, compared with the static grasping strategy, the dynamic grasping strategy based on collision makes the grasping of the target object more smooth, continuous, and intelligent.

[0058] In some embodiments, the process of controlling the robot to collide with the target object can be more specifically described.Figure 2 An exemplary flowchart of a process S101 of controlling the collision of the robot with the target object according to an embodiment of the present disclosure is shown.

[0059] Referring to Figure 2 First, in step S1011, the motion state information of the target object at a collision initial time is determined.

[0060] The collision initial time refers to the time when the robot starts to collide with the target object. The collision initial time may, for example, be pre-set by the system or selected by the user according to actual needs.

[0061] The motion state information of the target object at the collision initial time includes, for example, one or more of the following: the position of the center of mass motion of the target object at the collision initial time, the velocity of the center of mass motion, the attitude angle, and the angular velocity of the attitude.

[0062] For example, the motion state information of the target object at the collision initial time can be determined according to the motion state information of the target object before the collision. For example, if the target object is in a stationary state before the collision, the motion state information (position of the center of mass motion, velocity of the center of mass motion, attitude angle, and angular velocity of the attitude) of the target object before the collision is the motion state information of the target object at the collision initial time. If the target object is in a motion state (e.g., uniform speed, uniform acceleration, variable acceleration) before the collision, the motion state information of the target object at the collision initial time can be calculated based on the determined collision initial time and the motion state information of the target object before the collision.

[0063] Subsequently, in step S1012, the desired motion state information of the target object at a collision end time and the desired motion state information of the robot at the collision end time are obtained.

[0064] The collision end time refers to the time when the robot completes the collision with the target object. The desired motion state information of the target object at the collision end time and the desired motion state information of the robot at the collision end time may, for example, be pre-set by the system or selected and determined by the user.

[0065] It should be understood that the desired motion information of the target object and the robot at the end time of the collision can be determined according to the desired post-collision motion type of the target object, the attribute parameter information of the target object (mass, stiffness, etc. of the object), and the motion state information of the target object at the initial time of the collision. For example, if it is understood based on the attribute parameter information of the target object that the stiffness of the target object is large, the desired post-collision motion state of the target object can be set to a separation type motion type, and the desired motion information of the target object at the end time of the collision and the desired motion information of the robot at the end time of the collision can be set based on the motion state information of the target object at the initial time of the collision and the separation type motion type, so that the desired separation type motion type can be achieved.

[0066] It should be understood that the desired motion state information of the target object at the end time of the collision includes one or more of the desired velocity of the collision point of the target object at the end time of the collision, and the desired center of mass motion position, the desired center of mass motion velocity, the desired attitude angle, and the desired attitude angular velocity of the target object at the end time of the collision.

[0067] The desired motion state information of the robot at the end time of the collision includes one or more of the desired center of mass motion position, the desired center of mass motion velocity, the desired attitude angle, and the desired attitude angular velocity of the robot at the end time of the collision.

[0068] Thereafter, in step S1013, based on the motion state information of the target object at the initial time of the collision, the desired motion state information of the target object at the end time of the collision, and the desired motion state information of the robot at the end time of the collision, the target motion state information of the robot at the initial time of the collision is determined based on a collision motion model.

[0069] The collision motion model refers to a motion model established based on the dynamic characteristics of the robot and the target object during the collision, which aims to characterize the relationship between the motion state of the robot and the target object before the collision and the motion state of the robot and the target object after the collision.

[0070] It should be appreciated that determining the target motion state information of the robot manipulator at the initial time of the collision based on the collision motion model includes: determining the expected collision impulse of the target object based on the motion state information of the target object at the initial time of the collision, the expected motion state information of the target object at the end time of the collision; determining the expected motion velocity of the collision point of the target object at the initial time of the collision based on the expected collision impulse of the target object, the expected motion state information of the target object at the end time of the collision, and the expected motion state information of the robot manipulator at the end time of the collision; and determining the target motion state information of the robot manipulator at the initial time of the collision based on the expected motion velocity of the collision point of the target object at the initial time of the collision and the motion state information of the target object at the initial time of the collision.

[0071] However, it should be appreciated that the above only gives an exemplary method of determining the target motion state information of the robot manipulator at the initial time of the collision. According to actual needs, other ways can also be used to determine the target motion state information of the robot manipulator at the initial time of the collision.

[0072] Finally, in step S1014, the robot manipulator is controlled to collide with the target object based on the target motion state information of the robot manipulator at the initial time of the collision.

[0073] For example, the robot manipulator can be adjusted based on the pre-collision motion state information (such as the motion velocity, motion position, motion acceleration, attitude angle, attitude angular velocity, etc. of the robot manipulator before the collision) and the target motion state information of the robot manipulator at the initial time of the collision, so that the robot manipulator collides with the target object and has the target motion state information at the initial time of the collision.

[0074] Based on the above, in the present application, in the process of controlling the robot manipulator to collide with the target object, the motion state information of the target object at the initial time of the collision is determined, the expected motion state information of the target object at the end time of the collision and the expected motion state information of the robot manipulator at the end time of the collision are obtained, and the target motion state information of the robot manipulator at the initial time of the collision is determined based on the above information according to the collision motion model, and the robot manipulator is controlled to collide with the target object. This makes it possible to solve the motion state that the robot manipulator should have at the initial time of the collision based on the motion state of the target object at the initial time of the collision, the expected motion state at the end time of the collision, and the expected motion state of the robot manipulator at the end time of the collision according to the collision motion model, so as to accurately and reliably control the collision process of the robot manipulator and the target object.

[0075] In some embodiments, the motion state information of the target object at the initial time of the collision includes the motion position, motion velocity, attitude angle, and attitude angular velocity of the center of mass of the target object at the initial time of the collision.

[0076] The desired motion state information of the target object at the end time of the collision includes: a desired velocity of a collision point of the target object at the end time of the collision, and a desired center-of-mass motion position, a desired center-of-mass motion velocity, a desired attitude angle, and a desired attitude angular velocity of the target object at the end time of the collision.

[0077] The desired motion state information of the robot at the end time of the collision includes: a desired center-of-mass motion position, a desired center-of-mass motion velocity, a desired attitude angle, and a desired attitude angular velocity of the robot at the end time of the collision.

[0078] The target motion state information of the robot at the initial time of the collision includes: a target position, a target velocity, a target attitude angle, and a target attitude angular velocity of the robot at the initial time of the collision.

[0079] By setting the specific composition of the motion state information of the target object at the initial time of the collision, the desired motion state information of the target object at the end time of the collision, the desired motion state information of the robot at the end time of the collision, and the target motion state information of the robot at the initial time of the collision, the motion state of the robot and the target object at the initial time and the end time of the collision motion can be well reflected based on the information, so that the target motion state information of the robot at the initial time of the collision can be accurately determined, and the collision process can be reliably controlled.

[0080] In some embodiments, the process of determining the target motion state information of the robot at the initial time of the collision based on the motion state information of the target object at the initial time of the collision, the desired motion state information of the target object at the end time of the collision, and the desired motion state information of the robot at the end time of the collision based on the collision motion model can be described more specifically. Figure 3A An exemplary flowchart of the process S1013 of determining the target motion state information of the robot at the initial time of the collision based on the collision motion model according to an embodiment of the present disclosure is shown.

[0081] Referring to Figure 3A , first, in step S1013-1, the desired collision impulse of the target object is determined based on the motion state information of the target object at the initial time of the collision and the desired motion state information of the target object at the end time of the collision.

[0082] The desired collision impulse refers to the impulse value that the target object is expected to accumulate during the collision process in order for the target object to have the desired motion state information at the end time of the collision.

[0083] Thereafter, in step S1013-2, based on the desired impact impulse of the target object, the desired motion state information of the target object at the end time of the impact, and the desired motion state information of the robot at the end time of the impact, the desired motion speed of the impact point of the target object at the initial time of the impact is determined.

[0084] Finally, in step S1013-3, based on the desired motion speed of the impact point of the target object at the initial time of the impact and the motion state information of the target object at the initial time of the impact, the target motion state information of the robot at the initial time of the impact is determined.

[0085] Next, the above process will be described in more detail in connection with a specific impact motion model. Figure 3B A schematic diagram showing the impact process of the robot and the target object is shown. Referring to Figure 3B where the target object M, the robot S, the impact point C of the target object and the robot, and the motion positions of the center of mass of the target object and the robot in the XY plane at the initial time of the impact (x0, y0) and at the end time of the impact (x1, y1) are shown. For example, the attitude angular velocities of the robot and the target object at the initial time of the impact are ω0x, ω0y, and ω0z, respectively. s0 s0

[0086] At this time, for example, based on the dynamic characteristics of the robot and the target object during the impact process, the impact models of the robot and the target object can be constructed as follows:

[0087] First, the speed of the impact point C of the robot and the target object at the initial time of the impact can be determined according to the motion state information of the robot and the target object at the initial time of the impact as follows:

[0088]

[0089]

[0090] where vcyis the center of mass motion speed component of the impact point C in the y-axis direction on the XY plane, vcyis the center of mass motion speed component of the impact point C in the y-axis direction on the XY plane, vcyis the center of mass motion speed component of the impact point C in the y-axis direction on the XY plane; vcyis the center of mass motion speed component of the impact point C in the y-axis direction on the XY plane, ​​a center-of-mass motion velocity component of the target object along the y-axis in the XY plane at the initial time (T0) of the collision, a center-of-mass motion velocity component of the robot along the x-axis in the XY plane at the initial time (T0) of the collision, a center-of-mass motion velocity component of the robot along the y-axis in the XY plane at the initial time (T0) of the collision. The meanings of the remaining parameters are as described above.

[0091] In addition, considering that the collision process undergoes a compression and recovery phase, during which it is assumed that the target object does not move, and the velocities of the target object and the robot change with the accumulated impulse P, it can be seen that the effect of the impulse on the velocity of the target object is:

[0092]

[0093]

[0094]

[0095] where m is the equivalent collision mass of the target object, I is the equivalent moment of inertia of the target object, and m and I are, for example, preset values. a center-of-mass motion velocity component of the target object along the x-axis in the XY plane at the end time (Tt) of the collision, a center-of-mass motion velocity component of the target object along the y-axis in the XY plane at the end time (Tt) of the collision. x t a transverse center-of-mass motion position of the target object in the x-axis direction in the XY plane at the end time (Tt) of the collision; y t a longitudinal center-of-mass motion position of the target object in the y-axis direction in the XY plane at the end time (Tt) of the collision, (x t , y t ) being the center-of-mass motion position information of the target object in the XY plane. P x a transverse impulse of the target object in the x direction in the collision, P y a longitudinal impulse of the target object in the y direction in the collision, the meanings of the remaining parameters being as described above.

[0096] And as the collision motion proceeds, the impulse P generated in the target object will change the velocity of the collision point, according to dynamics and kinematics, the expression of the velocity of the collision point at the end time of the collision is as follows:

[0097]

[0098]

[0099] wherein, is the velocity of the collision point of the target object along the y-axis direction at the end time of the collision, is the velocity of the collision point of the target object along the x-axis direction at the end time of the collision. And wherein:

[0100]

[0101]

[0102]

[0103] wherein, x st is the mass center motion position of the robot along the x-axis direction at the end time of the collision, y st is the mass center motion position of the robot along the y-axis direction at the end time of the collision.

[0104] Based on the above formulas 1) -7), the collision motion model of the robot and the target object is obtained. At this time, based on the collision motion model, the target motion state information of the robot at the initial time of the collision is determined, for example:

[0105] First, based on the motion state information of the target object at the initial time of the collision, the expected motion state information of the target object at the end time of the collision, according to formulas 3) -5) in the collision motion model, the expected collision impulse of the target object is determined Specifically:

[0106]

[0107]

[0108]

[0109] wherein, the expected motion state information of the target object at the end time of the collision includes: the expected mass center motion position of the target object at the end time of the collision the expected mass center motion velocity of the target object at the end time of the collision The remaining parameters in the formula are as described before.

[0110] The expected collision impulse of the target object is obtained After that, for example, according to formulas 6) -7) in the collision motion model, based on the expected collision impulse of the target object, the expected motion state information of the target object at the end time of the collision, the expected motion state information of the robot at the end time of the collision, the expected motion velocity of the collision point of the target object at the initial time of the collision is determined Specifically:

[0111]

[0112]

[0113] and wherein the desired motion state information of the robot at the end time of the collision further comprises: a desired motion velocity of the collision point of the robot at the end time of the collision the remaining parameter meanings are as described above, and the parameters B1, B2, B3 are calculated according to the desired motion state information (for example, the desired mass center motion position) of the robot at the end time of the collision and the desired motion state information (for example, the desired mass center motion position) of the target object at the end time of the collision.

[0114] Finally, based on the calculated desired motion velocity of the collision point of the target object at the initial time of the collision The motion state information (position, velocity, angular velocity, angle) of the target object at the initial time of the collision can be used to determine the target motion state information of the robot at the initial time of the collision based on the aforementioned formulas 1)-2), specifically:

[0115]

[0116]

[0117] and wherein, is the aforementioned calculated desired motion velocity of the collision point at the initial time, and the remaining parameter meanings are as described above, whereby the target position, target velocity, target attitude angle, and target attitude angular velocity of the robot at the initial time of the collision can be calculated, i.e., the target motion state information of the robot at the initial time of the collision can be obtained.

[0118] It should be understood that the above only gives an exemplary collision motion model of a robot and a target object. According to actual needs, collision motion models with other component configurations can also be constructed. The embodiments of the present disclosure are not limited by the specific components of the collision motion model.

[0119] Based on the above, in the present application, when calculating the target motion state information of the robot at the initial time of the collision, the desired collision impulse of the target object is first determined based on the collision motion model; the desired motion velocity of the collision point of the target object at the initial time of the collision is determined based on the desired collision impulse of the target object; and the target motion state information of the robot at the initial time of the collision is determined based on the desired motion velocity of the collision point of the target object at the initial time of the collision, so that the determination of the target motion state information of the robot at the initial time of the collision can be accurately and reliably realized while well satisfying the kinematic constraints of the collision.

[0120] In some embodiments, the process S102 of determining the post-collision movement type of the target object may, for example, include: first, obtaining the movement state information of the collision point of the target object at the end time of the collision and the movement state information of the robot at the end time of the collision.

[0121] The movement state information of the collision point of the target object at the end time of the collision may, for example, include the movement speed of the collision point at the end time of the movement. The movement state information of the robot at the end time of the collision may, for example, include the centroid movement speed of the robot. However, it should be understood that other movement state information may also be obtained according to actual needs.

[0122] Thereafter, the movement state information of the collision point of the target object at the end time of the collision is compared with the movement state information of the robot at the end time of the collision, and the post-collision movement type of the target object is determined based on the comparison result.

[0123] For example, the movement speed of the collision point of the target object at the end time of the collision may be compared with the centroid movement speed of the robot at the end time of the collision. If the movement speed of the collision point of the target object at the end time of the collision is greater than the centroid movement speed of the robot at the end time of the collision, the post-collision movement type of the target object may be determined as the escape type. If the movement speed of the collision point of the target object at the end time of the collision is less than or equal to the centroid movement speed of the robot at the end time of the collision, the post-collision movement type of the target object may be determined as the non-escape type.

[0124] Based on the above, in the present application, by comparing the movement state information of the collision point of the target object and the robot at the end time of the collision, the post-collision movement type of the target object can be determined in a simple and convenient manner, thereby facilitating subsequent implementation of different operations based on the post-collision movement type to grasp the target object.

[0125] In some embodiments, the process S103 of controlling the robot to grasp the target object based on the post-collision movement type of the target object may, for example, be more specifically illustrated. Figure 4 An exemplary flowchart of the process S103 of controlling the robot to grasp the target object according to an embodiment of the present disclosure is shown.

[0126] Referring to Figure 4 , first, in step S1031, the robot is controlled to grasp the target object based on the post-collision movement type of the target object, so that the target object is located on the robot and the pose of the robot is consistent with that of the target object.

[0127] Specifically, for example, the robot hand comprises a palm base and a plurality of finger portions extending outwardly from the palm base, the grasped state is that the target object is located on the palm base of the robot hand and in contact with the palm base, and the target object and the palm base of the robot hand have the same attitude angle.

[0128] It should be appreciated that, as will be described in detail below, the robot hand can be controlled to grasp the target object by different motion control models and control modes according to different post-collision motion types and actual needs, and embodiments of the present disclosure are not limited by the specific manner in which the robot hand grasps the target object.

[0129] After the grasping step is performed, in step S1032, the robot hand is controlled to perform a gripping operation to grasp the target object. The gripping operation refers to an operation of controlling the end of the robot hand to close to grasp the target object.

[0130] Based on this, in the present application, the robot hand is controlled to grasp the target object based on the post-collision motion type of the target object, and then the robot hand is controlled to perform a gripping operation to grasp the object, so that the dynamic grasping process of the target object can be realized in a stable and reliable manner.

[0131] Next, the process of controlling the robot hand to grasp the target object under the non-detached motion type and the detached motion will be described in more detail.

[0132] It should be appreciated that, regardless of the post-collision motion type of the target object, the motion of the target object in the target object coordinate system satisfies the following dynamic equation:

[0133]

[0134] In the formula, M(x) is the inertia matrix of the target object, is the centrifugal force item matrix of the target object, and G(x) is the gravity matrix of the target object. M is the equivalent contact force and torque vector of the object. is the acceleration, velocity, and displacement vector of the object. Based on different post-collision motion types, the expression of F M is different.

[0135] wherein the equivalent contact force F M can be expressed as shown below:

[0136] F M =G r f c 16)

[0137] wherein: G ris the grasp matrix, fcis the equivalent force of the manipulator acting on the target object at the contact point. And wherein the relevant parameters can be expressed as follows:

[0138]

[0139] f c ∈ FC, FC = FC1x FC2x... FC K

[0140]

[0141] And wherein FC i is the friction cone at the ith contact point.

[0142] Specifically, in some embodiments, the target post-collision motion type is a separation motion type. At this time, the target object flies out after colliding with the manipulator, and at this time, the target object only moves (rotates) under the action of gravity, Figure 5A A schematic diagram of an object in a separation motion type is shown according to an embodiment of the present disclosure. Referring to Figure 5A At the end of the collision, the target object separates from the manipulator, and starts to rotate only under the action of gravity G with an initial attitude angle a with the horizontal line. At this time, the pose and position of the robot manipulator should be adjusted so that the manipulator completes the grasping of the target object at the optimal grasping angle a0, and then controls the closing of the finger part to grasp the target object.

[0143] Figure 5B An exemplary flowchart of the process S1031A of controlling the manipulator to grasp the target object when the target post-collision motion type is a separation motion type is shown according to an embodiment of the present disclosure.

[0144] Referring to Figure 5B First, in step S1031A-1, the post-collision motion state information of the target object is obtained, and based on the post-collision motion state information of the target object, the post-collision motion trajectory and attitude estimation of the target object are generated according to the separation motion model.

[0145] The post-collision motion state information of the target object refers to information reflecting the motion state characteristics of the target object after the collision. The post-collision motion state information, for example, includes at least one of the center of mass motion speed, the center of mass motion position, the attitude angle, and the attitude angular velocity of the target object at the end of the collision and within a preset time period after the collision, or can also be the motion trajectory of the target object at the end of the collision and within a preset time period after the collision. It should be understood that the post-collision motion state information of the target object can also be obtained according to actual needs.

[0146] The disengagement motion model refers to a model for reflecting the dynamic characteristics of the target object in the disengagement motion process. Specifically, when the target object is in the disengagement flight stage, the equivalent contact force received is only gravity, and since the target object is not in contact with the robot hand, the force at each contact point is 0, that is, at this time: F M =G r f c =0. Thus, for example, the disengagement motion model of the target object under the disengagement motion type can be obtained as follows:

[0147]

[0148] At this time, based on the disengagement motion model and the post-collision motion state information of the target object, the post-collision motion trajectory and attitude estimation of the target object can be calculated based on a preset algorithm.

[0149] The post-collision motion trajectory and attitude estimation refer to the estimation of the motion trajectory of the target object from the end time of the collision to the end time of the disengagement motion when it falls to the horizontal plane (such as the ground or the desktop), and the estimation of the attitude angle and attitude angular velocity of the target object in the estimated motion trajectory.

[0150] Thereafter, in step S1031A-2, the desired grasping motion state information of the robot hand is determined based on the post-collision motion trajectory and attitude estimation of the target object.

[0151] The desired grasping motion state information includes, for example, desired grasping position information, desired grasping speed information, and desired grasping attitude information. The desired grasping motion state information of the robot hand refers to the motion state that the robot hand is expected to have in order to achieve good grasping of the target object. The desired grasping position information refers to the desired position of the robot when it contacts the target object; the desired grasping speed information refers to the desired speed of the robot when it contacts the target object; and the desired grasping attitude information refers to the attitude angle and angular velocity data that the robot is expected to have in the case of good grasping of the target object (both have the same attitude).

[0152] For example, the desired grasping motion state information of the robot hand can be determined comprehensively according to the post-collision motion trajectory and attitude estimation of the target object and the motion state information (such as the current motion speed, motion position, attitude angle, etc.) of the robot hand itself after the collision.

[0153] Thereafter, in step S1031A-3, the robot hand is controlled to grasp the target object based on the desired grasping motion state information of the robot hand.

[0154] The process of controlling the robot to pick up the target object can be described more specifically. For example, based on the desired pick-up position information, the desired pick-up speed information, the motion path of the robot can be planned so that the robot can travel to the desired position with the desired speed via the shortest motion path in the shortest time and with the smallest range of speed adjustment (i.e. the smallest change in acceleration) to achieve contact with the target object (intercept the target object); in addition, based on the desired pick-up posture information, the robot is controlled so that the posture of the robot at the pick-up position is consistent with the target object (in the desired posture) to well pick up the target object.

[0155] Specifically, Figure 5C A schematic diagram of a robot picking up a target object is shown according to an embodiment of the present disclosure. Referring to Figure 5C where r m is the vector from the robot coordinate origin to the robot's arm, and rc is the vector between the robot system coordinate origin and the desired pick-up point of the target object. The boundary condition for the robot to successfully contact and intercept the target object after the object leaves the flight is that the position and speed of the robot's arm need to be consistent with the target object at the desired pick-up position, and the time, speed and acceleration weighted optimization needs to be maintained throughout the process, so that the entire process can be represented as the following optimization process:

[0156]

[0157] r m (t f )=r C (t f )

[0158]

[0159] In the above formula, w1 and w2 are the weight coefficients of the speed term and the acceleration term of the robot's arm, respectively. t f is the termination time of the robot's arm intercepting the target object. T is the time length from the end of the collision to the time when the robot's arm contacts the target object at the desired pick-up position.

[0160] Based on the above, after completing the pick-up of the target object, the target object will have the same posture as the robot's arm and be located on the robot's arm.

[0161] Based on the above, in the present application, in the case that the target object is in the disengagement motion type after the collision, the post-collision motion trajectory and attitude estimation of the target object are generated based on the post-collision motion state information of the target object and the disengagement motion model, the expected grasping motion state information of the robot hand is determined based on the motion trajectory and attitude estimation, and the robot hand is controlled to grasp the target object, so that the reliable and accurate grasping of the target object can be realized while satisfying the kinetic constraint of the disengagement motion type of the target object, which is beneficial to subsequent grasping of the target object.

[0162] In some embodiments, in the case that the target post-collision motion type is the non-disengagement motion type, at this time, the target object does not completely disengage from the robot hand after the collision, and performs complex motion including rotation under the action of the force Fn of the robot hand and the gravity G. Figure 6A A schematic diagram of the motion process of the target object in the non-disengagement motion type is shown according to an embodiment of the present disclosure, wherein the x-axis and y-axis of the XY plane are also schematically shown.

[0163] In the non-disengagement motion type, there are two contact points between the robot hand and the target object, so that the operability of the robot hand to the target object is reduced. At this time, the operation of the robot hand to the target object is mainly to control the rotational motion of the target object. At this time, the attitude of the robot hand and the contact force acting on the target object should be adjusted, so that the robot hand completes the grasping of the target object at the optimal grasping angle a0, and then the finger of the robot hand is controlled to be closed to grasp the target object.

[0164] At this time, the process of controlling the robot hand to grasp the target object can be explained more specifically, for example. Figure 6B An exemplary flowchart of the process S1031B of controlling the robot hand to grasp the target object in the case that the target object is in the non-disengagement motion type after the collision according to an embodiment of the present disclosure is shown.

[0165] Referring to Figure 6B Firstly, in step S1031B-1, the post-collision motion state information of the target object is obtained.

[0166] As described above, the post-collision motion state information of the target object refers to information reflecting the motion state characteristics of the target object after the collision. The post-collision motion state information includes, for example, at least one of the center of mass motion speed, the center of mass motion position, the attitude angle, and the attitude angular velocity of the target object within a preset time period after the end of the collision, or the motion trajectory of the target object within the preset time period after the end of the collision. It should be understood that the post-collision motion state information of the target object can also be obtained according to actual needs.

[0167] Afterwards, in step S1013B-2, the expected post-collision motion state information of the target object corresponding to the non-detached motion type is obtained.

[0168] The expected post-collision motion state information of the target object corresponding to the non-detached motion type refers to the expected motion state of the target object under the non-detached motion type. The expected post-collision motion state information includes, for example, an expected motion trajectory, an expected motion posture, etc. The embodiments of the present disclosure are not limited by the specific composition of the expected post-collision motion information.

[0169] It should be understood that the above steps S1013B-1 and S1013B-2 can be executed in sequence, in reverse order, or simultaneously. The embodiments of the present disclosure are not limited by the execution order of the steps S1013B-1 and S1013B-2.

[0170] Afterwards, in step S1013B-3, based on the post-collision motion state information of the target object and the expected post-collision motion state information of the target object corresponding to the non-detached motion type, the expected grasping information of the robot acting on the target object is determined according to the non-detached motion model, and the expected grasping information includes an expected contact force and an expected contact posture.

[0171] The expected grasping information refers to the expected motion information of the robot for achieving good grasping of the target object under the non-detached motion type. The expected contact force refers to the contact force expected to be obtained by the target object from the robot, and the expected contact posture refers to the expected posture state of the robot when the target object is located on the robot and has the same posture as the robot.

[0172] At this time, the non-detached motion model of the target object can be, for example,

[0173]

[0174] wherein, is the expected center-of-mass motion acceleration of the target object, is the expected center-of-mass motion velocity of the target object, is the expected contact force of the target object. Thus, the expected contact force of the robot can be calculated

[0175] Afterwards, the expected contact posture of the robot can be further determined based on, for example, the expected contact force of the robot and the expected post-collision motion state information of the target object corresponding to the non-detached motion type.

[0176] After obtaining the expected grasping information, in step S1013B-4, the current grasping information of the robot acting on the target object is obtained, and the robot is controlled to grasp the target object based on the current grasping information and the expected grasping information.

[0177] For example, the current grasping information includes the current contact force of the robot acting on the target object, the current pose angle of the robot. The control process can be, for example, adjusting the current pose angle of the robot to be equal to the expected pose angle, and adjusting the current contact force of the robot to be equal to the expected contact force.

[0178] Based on the above, in this application, when the target object is in the non-decoupling motion type after the collision, by obtaining the expected post-collision motion state information of the target object, based on the expected post-collision motion state information of the target object, the expected grasping information of the robot acting on the target object is determined according to the non-decoupling motion model, and the control of the grasping process of the robot is realized, so that the grasping of the target object can be reliably and efficiently realized.

[0179] In some embodiments, after the robot grasps the target object, the robot is further configured to perform a grasping holding process (also known as a soft palm grasping) before the robot grasps the target object. Specifically, in this process, the robot's hand (for example, the palm base of the robot) will maintain the same pose as the target object, and the target object will always be on the robot and will not slip off.

[0180] Specifically, when the robot grasps the target object, the robot needs to adjust the pose to be completely consistent with the target object, and after the robot grasps the target object, the robot needs to maintain acceleration in the tilting direction, so as to increase the support force between the target object and the robot through the inertial force, and then maintain the friction force between the target object and the robot.

[0181] At this time, the contact between the robot (for example, the palm base of the robot) and the target object is a contact with four contact points. In order to prevent the object from falling, the contact force between the object and the robot needs to satisfy the following conditions:

[0182]

[0183] Where P and Q are the respective weights. The target object is the expected equivalent contact force of the object, F M is the actual equivalent contact force of the target object.

[0184] In some embodiments, the robot hand comprises a palm base and a plurality of finger portions extending outwardly from the palm base, for example, three finger portions or five finger portions. Embodiments of the present disclosure are not limited by the specific number of finger portions or their configurations.

[0185] At this time, the control of the robot hand to perform the gripping operation to grasp the target object comprises: controlling the plurality of finger portions to perform a gripping movement of approaching each other; acquiring contact force data of the end of at least one of the plurality of finger portions; and stopping the gripping movement when the contact force data is greater than or equal to a preset contact force threshold.

[0186] For example, the contact force data of the end of only one finger portion can be acquired, or the contact force data of the ends of multiple finger portions can be acquired. Embodiments of the present disclosure are not limited by the number of finger portion ends collected.

[0187] For example, the contact force data of the end of the finger portion can be acquired based on a force sensor arranged on each finger portion. Alternatively, other ways of acquiring the contact force data of the end of the finger portion can be used.

[0188] The contact force data represents the contact force between the finger portion and the target object. The preset contact force threshold, for example, is the expected contact force data between the finger portion and the target object in the case that the finger portion grasps the target object well; it can be set according to the object attribute parameters or can be preset by the system. Embodiments of the present disclosure are not limited by the specific data of the preset contact force threshold or the setting method.

[0189] In the present application, during the gripping operation, the contact force data of the end of the finger portion acting on the target object is collected in real time, and the gripping operation is stopped when the collected contact force data is greater than or equal to the preset contact force threshold. On the one hand, this enables good control of the gripping process, thereby enabling good grasping of the target object; on the other hand, further increase in contact force can be stopped in time after grasping, preventing additional pressure on the target object or damage to the target object due to continuous pressure.

[0190] In some embodiments, the current grasping information comprises: a current contact force of the robot hand acting on the target object.

[0191] And acquiring the current contact force of the robot hand acting on the target object comprises: acquiring current motion state information of the target object; determining an external force acting on the target object based on the current motion state information of the target object based on a dynamics equation; and determining the current contact force based on the external force.

[0192] Specifically, the external force applied to the target object can be determined based on the current motion state information of the target object, for example, by the aforementioned dynamic equation 15), and the current contact force applied by the robot to the target object can be determined based on the external force and the different motion states and force conditions of the target object.

[0193] Based on the above, in the present application, the motion state information of the target object is detected, the external force applied to the target object is determined based on the motion state information via a dynamic equation, and the current contact force applied by the robot to the target object is determined based on the external force. Compared with the case of detecting the contact force by setting a force sensor on the robot, determining the contact force based on the motion state of the target object enables real-time determination of the current contact force in a simple and convenient manner, thereby facilitating the simplification of the component settings of the robot and improving the detection accuracy.

[0194] Next, the robot control method in the present application will be described in more detail in conjunction with specific embodiments. Figure 7 A schematic diagram showing the robot performing a preset motion is shown. Figure 8 A schematic diagram showing the robot grasping a target object during the performance of a preset motion is shown.

[0195] The robot, for example, includes a robot hand and a robot arm connected to the robot hand for controlling the motion of the robot hand, and the robot hand, for example, has a palm base and four finger-shaped parts extending outward from the palm base. Figure 7 The robot hand components of the robot are shown in FIG. Figure 7 The robot hand, for example, is configured to perform a preset motion from a starting point A to an ending point B, for example, the robot hand is planned to move uniformly along a specific trajectory from point A to point B.

[0196] If there is a target object M on the preset motion trajectory of the robot hand (at this time, the target object is in a static state, for example), the robot will perform a collision-based dynamic grasping process according to the robot control method in the present application. Specifically, when the robot detects a target object M on or near the preset motion trajectory, the robot can perform the process of the aforementioned step S101, for example, to control the robot hand to collide with the target object. The control process, for example, can be as previously described in conjunction with Figure 2The target object is collided with the robot arm based on the target motion state information of the robot arm at the initial time of the collision. The robot arm collides with the target object based on the target motion state information of the robot arm at the initial time of the collision.

[0197] Thus, the robot arm collides with the target object based on the target motion state information of the robot arm at the initial time of the collision. Figure 8 The schematic diagram of the collision process between the robot arm and the target object is shown in FIG. 1. After the collision, the robot further determines the post-collision motion type of the target object based on the process of the foregoing step S102. The post-collision motion type of the target object includes a disengagement motion type and a non-disengagement motion type. The process of determining the post-collision motion type of the target object is, for example, specifically described in the foregoing reference Figure 3A and Figure 3B which will not be repeated here. Figure 8 The motion state of the target object under the disengagement motion type and the non-disengagement motion type is shown in FIG. 2. Thereafter, the robot controls the robot arm to grasp the target object based on the post-collision motion type of the target object, for example, by performing the process of step S103. Specifically, the robot, for example, first controls the robot arm to pick up the target object based on the post-collision motion type of the target object, so that the target object is located on the robot arm and the pose of the robot arm is consistent with that of the target object. The picking-up process under different post-collision motion types is, for example, specifically described in the foregoing reference Figures 5A-5C , Figures 6A-6B which will not be repeated here. Figure 8 The state of finally achieving the picking-up of the target object under the two different post-collision motion types is shown in FIG. 3. Thereafter, the robot controls the robot arm to perform a gripping operation to grasp the target object, and the state of grasping the target object is shown in FIG. 4. Figure 8

[0198] ​Thus, when there is a target object in the motion trajectory of the robot, the target object is given an initial speed by colliding the target object with the robot in a target motion state, and when the subsequent picking and gripping operations are completed to grasp the target object, the speed of the robot does not need to be reduced, that is, compared with the case where there is no target object, when there is a target object, the motion trajectory of the robot can only change in the robot pose, and there is no change in the robot motion coordinate position, and the robot can be made to not perform a deceleration process. Based on this, compared with the traditional static and quasi-static grasping process, such a dynamic grasping process based on collision is more efficient, fast, and smooth in trajectory.

[0199] According to another aspect of the present disclosure, a robot motion control system is provided. Figure 9 An exemplary block diagram of a self-balancing robot control system 600 according to an embodiment of the present disclosure is shown.

[0200] The robot includes a robot arm, and the system includes a collision control module 610, a post-collision motion type determination module 620, and a grasping module 630.

[0201] The collision control module 610 is configured to perform Figure 1 the process in step S101 to control the robot arm to collide with the target object.

[0202] It should be understood that the target object refers to an object that the robot aims to grasp. The target object may, for example, be in a static or motion state (e.g., a uniform motion state, a uniform acceleration motion state, or a variable acceleration motion state), and embodiments of the present disclosure are not limited by the motion state of the target object.

[0203] The post-collision motion type determination module 620 is configured to perform Figure 1 the process in step S102 to determine the post-collision motion type of the target object, the post-collision motion type of the target object including a disengagement motion type and a non-disengagement motion type.

[0204] The post-collision motion type of the target object refers to the motion type of the target object relative to the robot arm after the target object collides with the robot arm. The disengagement motion type refers to the target object flying away from the robot arm after colliding with the robot arm. The non-disengagement motion type refers to the target object remaining in contact with the robot arm at at least one contact point after colliding with the robot arm, and having the same motion state as the robot arm at the contact point (i.e., both are in a coordinated motion state at the contact point).

[0205] The grasping module 630 is configured to perform Figure 1In the process in step S103, the robot arm is controlled to grasp the target object based on the post-collision motion type of the target object.

[0206] For example, the robot arm can be first controlled to perform a picking-up step on the target object based on the post-collision motion type of the target object, so that the target object is in a picked-up state; and then the robot arm is controlled to perform a gripping operation to grasp the target object.

[0207] Based on the above, in the present application, by setting a dynamic grasping strategy, specifically, by first controlling the robot arm to collide with the target object, determining the post-collision motion type of the target object, and controlling the robot arm to grasp the target object based on the determined post-collision motion type of the target object, the grasping of the target object by the robot during the motion process is realized. Compared with the current target object grasping process based on a static grasping strategy, on the one hand, by adopting a dynamic grasping strategy based on collision, the influence of the grasping operation on the motion task of the robot is reduced, so that the robot can achieve good and reliable grasping of the target object without the need to slow down or stop the motion, so that the process of grasping the target object does not affect the motion task of the robot, which is conducive to the robot to balance the execution of the target object grasping and the motion task itself. On the other hand, compared with the static grasping strategy, the dynamic grasping strategy based on collision makes the grasping of the target object more smooth, continuous, and intelligent.

[0208] In some embodiments, the collision control module 610 includes an object motion state determination module 611, an expected motion state acquisition module 612, a robot arm collision information generation module 613, and a collision generation module 614.

[0209] The object motion state determination module 611 is configured to perform Figure 2 In the process of step S1011, the motion state information of the target object at the initial collision time is determined.

[0210] The initial collision time refers to the time when the robot arm starts to collide with the target object. According to actual needs, the initial collision time may, for example, be pre-set by the system or selected by the user.

[0211] The motion state information of the target object at the initial collision time includes, for example, one or more of the mass center motion position, the mass center motion velocity, the attitude angle, and the attitude angular velocity of the target object at the initial collision time.

[0212] The expected motion state acquisition module 612 is configured to perform Figure 2In the process of step S1012, the expected motion state information of the target object at the end time of the collision and the expected motion state information of the robot at the end time of the collision are obtained.

[0213] The end time of the collision refers to the time when the robot and the target object complete the collision. The expected motion state information of the target object at the end time of the collision and the expected motion state information of the robot at the end time of the collision can be pre-set by the system or selected and determined by the user.

[0214] The robot collision information generation module 613 is configured to perform Figure 2 In the process of step S1013, based on the motion state information of the target object at the initial time of the collision, the expected motion state information of the target object at the end time of the collision, and the expected motion state information of the robot at the end time of the collision, the target motion state information of the robot at the initial time of the collision is determined based on a collision motion model.

[0215] The collision motion model refers to a motion model established based on the dynamic characteristics of the robot and the target object during the collision process, which aims to represent the relationship between the motion state of the robot and the target object before the collision and the motion state of the robot and the target object after the collision.

[0216] The collision generation module 614 is configured to perform Figure 2 In the process of step S1014, based on the target motion state information of the robot at the initial time of the collision, the robot collides with the target object.

[0217] Based on the above, in the process of controlling the robot to collide with the target object in the present application, the motion state information of the target object at the initial time of the collision is determined, and the expected motion state information of the target object at the end time of the collision and the expected motion state information of the robot at the end time of the collision are obtained. Based on the above information, the target motion state information of the robot at the initial time of the collision is determined according to the collision motion model, and the robot is controlled to collide with the target object. This makes it possible to accurately and reliably control the collision process of the robot and the target object based on the motion state of the target object at the initial time of the collision, the expected motion state at the end time of the collision, and the expected motion state of the robot at the end time of the collision.

[0218] In some embodiments, the motion state information of the target object at the initial time of the collision includes the mass center motion position, the mass center motion velocity, the attitude angle, and the attitude angular velocity of the target object at the initial time of the collision.

[0219] The expected motion state information of the target object at the end of the collision includes: the expected velocity of the collision point of the target object at the end of the collision, and the expected position of the center of mass, the expected velocity of the center of mass, the expected attitude angle, and the expected attitude angular velocity of the target object at the end of the collision.

[0220] The expected motion state information of the manipulator at the moment of collision termination includes: the expected position of the manipulator's center of mass, the expected velocity of the center of mass, the expected attitude angle, and the expected attitude angular velocity at the moment of collision termination.

[0221] The target motion state information of the robotic arm at the initial moment of the collision includes: the target position, target velocity, target attitude angle, and target attitude angular velocity of the robotic arm at the initial moment of the collision.

[0222] By setting the specific composition of the motion state information of the target object at the initial moment of the collision, the expected motion state information of the target object at the end moment of the collision, the expected motion state information of the robot at the end moment of the collision, and the target motion state information of the robot at the initial moment of the collision, it is possible to reflect the motion state of the robot and the target object at the beginning and end of the collision motion based on this information. This is beneficial to accurately determine the target motion state information of the robot at the initial moment of the collision, thereby reliably controlling the collision process.

[0223] In some embodiments, the robot collision information generation module 613 includes: a collision impulse calculation module 6131, a collision point initial velocity determination module 6132, and a robot initial velocity determination module 6133.

[0224] The collision impulse calculation module 6131 is configured to execute Figure 3A In step S1013-1, the expected collision impulse of the target object is determined based on the motion state information of the target object at the initial moment of the collision and the expected motion state information of the target object at the end moment of the collision.

[0225] The desired collision impulse refers to the impulse value that the target object is expected to accumulate during the collision so that it has the desired motion state information at the end of the collision.

[0226] The initial velocity determination module 6132 at the collision point is configured to execute Figure 3A In step S1013-2, based on the expected collision impulse of the target object, the expected motion state information of the target object at the end of the collision, and the expected motion state information of the robot at the end of the collision, the expected motion velocity of the collision point of the target object at the initial moment of the collision is determined.

[0227] The initial velocity determination module 6133 of the robotic arm is configured to execute...Figure 3A In step S1013-3, based on the expected velocity of the collision point of the target object at the initial time of the collision and the motion state information of the target object at the initial time of the collision, the target motion state information of the robot at the initial time of the collision is determined.

[0228] Based on the above, in the present application, when calculating the target motion state information of the robot at the initial time of the collision, the expected collision impulse of the target object is first determined based on the collision motion model; the expected velocity of the collision point of the target object at the initial time of the collision is determined based on the expected collision impulse of the target object; and thus the target motion state information of the robot at the initial time of the collision is determined based on the expected velocity of the collision point of the target object at the initial time of the collision, so that the determination of the target motion state information of the robot at the initial time of the collision can be accurately and reliably realized under the premise of well meeting the kinematic constraints of the collision.

[0229] In some embodiments, the post-collision motion type determination module 620 includes a collision end motion state acquisition module 621, a motion state comparison module 622, and a motion type determination module 623.

[0230] The collision end motion state acquisition module 612 is configured to acquire the motion state information of the collision point of the target object at the end time of the collision and the motion state information of the robot at the end time of the collision.

[0231] The motion state information of the collision point of the target object at the end time of the collision includes, for example, the velocity of the collision point at the end time of the motion. The motion state information of the robot at the end time of the collision includes, for example, the velocity of the robot. However, it should be understood that other motion state information can also be acquired according to actual needs.

[0232] The motion state comparison module 622 is configured to compare the motion state information of the collision point of the target object at the end time of the collision with the motion state information of the robot at the end time of the collision.

[0233] The motion type determination module 623 is configured to determine the post-collision motion type of the target object based on the comparison result.

[0234] Based on the above, in the present application, by comparing the motion state information of the collision point of the target object and the robot at the end time of the collision, the post-collision motion type of the target object can be determined in a simple and convenient manner, thereby facilitating subsequent implementation of different operations based on the post-collision motion type to achieve the grasping of the target object.

[0235] In some embodiments, the grasping module 630 includes a grasping operation module 631 and a clamping operation module 632.

[0236] The picking operation module 631 is configured to control the robot arm to pick the target object based on the post-collision motion type of the target object, so that the target object is on the robot arm and the robot arm is consistent with the pose of the target object.

[0237] Specifically, for example, the robot arm includes a palm base and a plurality of finger-shaped parts extending outward from the palm base, the picking state refers to the target object being on the palm base of the robot arm and being in contact with the palm base, and the target object and the palm base of the robot arm have the same pose angle.

[0238] The gripping operation module 632 is configured to control the robot arm to perform a gripping operation to grasp the target object.

[0239] Based on this, in the present application, by controlling the robot arm to pick the target object based on the post-collision motion type of the target object, and then controlling the robot arm to perform a gripping operation to grasp the object, the dynamic grasping process of the target object can be realized in a stable and reliable manner.

[0240] In some embodiments, the robot control system can perform the method as described above, with the functions as described above.

[0241] According to another aspect of the present disclosure, a robot is provided. The robot includes a robot arm, and wherein the robot has a robot control system as described above, capable of performing a robot control method as described above, to realize a collision-based dynamic grasping function as described above.

[0242] In addition, the robot can also include a bus, a memory, a sensor component, a controller, a communication module, and an input and output device, etc.

[0243] The bus can be a circuit that interconnects the components of the robot and transmits communication information (e.g., control messages or data) among the components.

[0244] The sensor component can be used to perceive the physical world, for example, including a camera, an infrared sensor, an ultrasonic sensor, etc. In addition, the sensor component can also include devices for measuring the current running and motion state of the robot, such as a Hall sensor, a laser position sensor, or a strain force sensor, etc.

[0245] A controller is used to control the operation of the robot, for example in an artificial intelligence controlled manner. The controller includes, for example, a processing device. The processing device can include a microprocessor, a digital signal processor ("DSP"), an application-specific integrated circuit ("ASIC"), a field-programmable gate array ("FPGA"), a state machine, or another processor that is configured to process electrical signals received from a sensor line. Such a processing device can include programmable electronic devices such as PLCs, programmable interrupt controllers ("PICs"), programmable logic devices ("PLDs"), programmable read-only memories ("PROMs"), electronically programmable read-only memories ("EPROMs" or "EEPROMs"), and the like.

[0246] The communication module can be wired or wireless, for example, to connect with a network to facilitate communication with the physical world (e.g., a server). The communication module can be wireless and can include a wireless interface such as IEEE 802.11, Bluetooth, a wireless local area network ("WLAN") transceiver, or a radio interface for accessing a cellular telephone network (e.g., a transceiver / antenna for accessing a CDMA, GSM, UMTS, or other mobile communication network). In another example, the communication module can be wired and can include an interface such as Ethernet, USB, or IEEE 1394.

[0247] The input / output device can transmit commands or data input from a user or any other external device to one or more other components of the robot, or can output commands or data received from one or more other components of the robot to the user or other external device.

[0248] A plurality of robots can constitute a robot system to cooperatively accomplish a task, the plurality of robots being communicatively connected to a server and receiving cooperative robot instructions from the server.

[0249] The present application uses certain terms to describe the embodiments of the present application. As used in the description of the application and the appended claims, the terms "first / second embodiment", "an embodiment", and / or "some embodiments" mean that a certain feature, structure, or characteristic described is a part of at least one embodiment. It is also noted that the features, structures, or characteristics of one or more embodiments of the present application can be combined in any manner.

[0250] Moreover, those skilled in the art will appreciate that the aspects of the application can be practiced with a variety of data processing systems or apparatuses, including personal computers, laptop computers, microprocessors, microcontrollers, programmable logic devices or combinations thereof. The aspects of the application can also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a data communications network. In a distributed computing environment, program modules can be located in both local and remote computer storage media including memory storage devices.

[0251] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

[0252] The above description is that of current embodiments of the application, and is not to be taken in a limiting sense but is made merely for the purpose of providing a factual description of the inventive embodiments. While the present application has been described in detail with respect to its exemplary embodiments, it will be understood that modifications can be made without departing from the true scope of the application, which is set forth in the following claims.

Claims

1. A robot motion control method, the robot comprising a manipulator, and the method comprising: controlling the manipulator to collide with a target object; determining a post-collision motion type of the target object, the post-collision motion type of the target object comprising: a disengagement motion type and a non-disengagement motion type; based on the post-collision motion type of the target object, controlling the manipulator to grasp the target object so that the target object is on the manipulator and the pose of the manipulator is consistent with the target object; and controlling the manipulator to perform a gripping operation to grasp the target object. controlling the manipulator to collide with the target object comprises: determining motion state information of the target object at a collision initial time; obtaining desired motion state information of the target object at a collision end time, desired motion state information of the manipulator at the collision end time; determining target motion state information of the manipulator at the collision initial time based on a collision motion model based on the motion state information of the target object at the collision initial time, the desired motion state information of the target object at the collision end time, the desired motion state information of the manipulator at the collision end time; and controlling the manipulator to collide with the target object based on the target motion state information of the manipulator at the collision initial time. 3.The robot motion control method of claim 2, wherein the motion state information of the target object at the collision initial time comprises: a center of mass motion position, a center of mass motion velocity, an attitude angle, and an attitude angular velocity of the target object at the collision initial time; the desired motion state information of the target object at the collision end time comprises: a desired velocity of a collision point of the target object at the collision end time, and desired center of mass motion position, desired center of mass motion velocity, desired attitude angle, and desired attitude angular velocity of the target object at the collision end time; the desired motion state information of the manipulator at the collision end time comprises: desired center of mass motion position, desired center of mass motion velocity, desired attitude angle, and desired attitude angular velocity of the manipulator at the collision end time; and the target motion state information of the manipulator at the collision initial time comprises: target position, target velocity, target attitude angle, and target attitude angular velocity of the manipulator at the collision initial time. determining the target motion state information of the manipulator at the collision initial time based on the collision motion model based on the motion state information of the target object at the collision initial time, the desired motion state information of the target object at the collision end time, the desired motion state information of the manipulator at the collision end time comprises: determining desired collision impulse of the target object based on the motion state information of the target object at the collision initial time, the desired motion state information of the target object at the collision end time; and determining desired motion velocity of the collision point of the target object at the collision initial time based on the desired collision impulse of the target object, the desired motion state information of the target object at the collision end time, the desired motion state information of the manipulator at the collision end time. ​ 2. The robot motion control method according to claim 1, wherein, ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ 4. The robot motion control method according to claim 2, wherein, ​ ​ ​ determine target motion state information of the robot at the initial time of the collision based on the expected motion velocity of the collision point of the target object at the initial time of the collision and motion state information of the target object at the initial time of the collision.

5. The robot motion control method of claim 1, wherein, The determination of the post-collision motion type of the target object comprises: obtaining motion state information of the collision point of the target object at the end time of the collision and motion state information of the robot at the end time of the collision; comparing the motion state information of the collision point of the target object at the end time of the collision with the motion state information of the robot at the end time of the collision; determining the post-collision motion type of the target object based on a comparison result.

6. The robot motion control method of claim 1, wherein, In a case where the post-collision motion type of the target object is the non-detachment motion type, the control of the robot to pick up the target object based on the post-collision motion type of the target object comprises: obtaining post-collision motion state information of the target object; obtaining expected post-collision motion state information of the target object corresponding to the non-detachment motion type; determining expected pick-up information of the robot acting on the target object according to a non-detachment motion model based on the post-collision motion state information of the target object and the expected post-collision motion state information of the target object corresponding to the non-detachment motion type, the expected pick-up information comprising: expected contact force, expected contact position, and expected contact posture; obtaining current pick-up information of the robot acting on the target object, and controlling the robot to pick up the target object based on the current pick-up information and the expected pick-up information.

7. The robot motion control method of claim 1, wherein, In a case where the post-collision motion type of the target object is the detachment motion type, the control of the robot to pick up the target object based on the post-collision motion type of the target object comprises: obtaining post-collision motion state information of the target object, and generating post-collision motion trajectory and posture estimation of the target object according to a detachment motion model based on the post-collision motion state information of the target object; determining expected pick-up motion state information of the robot based on the motion trajectory and the posture estimation of the target object, the expected pick-up motion state information comprising: expected pick-up position information, expected pick-up velocity information, and expected pick-up posture information; controlling the robot to pick up the target object based on the expected pick-up motion state information of the robot.

8. The robot motion control method of claim 1, wherein, The robot comprises: a palm base and a plurality of finger-shaped parts extending outward from the palm base, and the control of the robot to perform a gripping operation to grasp the target object comprises: controlling the plurality of finger-shaped parts to perform a gripping motion of approaching each other; obtaining contact force data of an end of at least one of the plurality of finger-shaped parts; and stopping the gripping motion when the contact force data is greater than or equal to a preset contact force threshold.

9. The robot motion control method of claim 6, wherein, The current pick-up information comprises: current contact force of the robot acting on the target object, and the obtaining of the current contact force of the robot acting on the target object comprises: obtaining current motion state information of the target object; determining external force acting on the target object based on a dynamics equation based on the current motion state information of the target object. determine a current contact force based on the external force. 10.A robot motion control system, the robot comprising a manipulator, and the system comprising: a collision control module configured to control the manipulator to collide with a target object; a post-collision motion type determination module configured to determine a post-collision motion type of the target object, the post-collision motion type of the target object comprising: a disengagement motion type and a non-disengagement motion type; a grasping module configured to control the manipulator to grasp the target object based on the post-collision motion type of the target object, wherein the grasping module comprises: a picking operation module configured to control the manipulator to pick the target object based on the post-collision motion type of the target object, so that the target object is on the manipulator and the manipulator is consistent with a pose of the target object; and a clenching operation module configured to control the manipulator to perform a clenching operation to grasp the target object.

11. The robotic motion control system of claim 10, wherein, the collision control module comprises: an object motion state determination module configured to determine motion state information of the target object at a collision initial time; an expected motion state acquisition module configured to acquire expected motion state information of the target object at a collision end time, expected motion state information of the manipulator at the collision end time; a manipulator collision information generation module configured to determine target motion state information of the manipulator at the collision initial time based on a collision motion model based on the motion state information of the target object at the collision initial time, the expected motion state information of the target object at the collision end time, and the expected motion state information of the manipulator at the collision end time; a collision generation module configured to control the manipulator to collide with the target object based on the target motion state information of the manipulator at the collision initial time. 12.The robot motion control system of claim 11, wherein the motion state information of the target object at the collision initial time comprises: a center-of-mass motion position, a center-of-mass motion velocity, a pose angle, and a pose angular velocity of the target object at the collision initial time; the expected motion state information of the target object at the collision end time comprises: an expected velocity of a collision point of the target object at the collision end time, and an expected center-of-mass motion position, an expected center-of-mass motion velocity, an expected pose angle, and an expected pose angular velocity of the target object at the collision end time; the expected motion state information of the manipulator at the collision end time comprises: an expected center-of-mass motion position, an expected center-of-mass motion velocity, an expected pose angle, and an expected pose angular velocity of the manipulator at the collision end time; the target motion state information of the manipulator at the collision initial time comprises: a target position, a target velocity, a target pose angle, and a target pose angular velocity of the manipulator at the collision initial time.

13. The robotic motion control system of claim 11, wherein, the manipulator collision information generation module comprises: a collision impulse calculation module configured to determine an expected collision impulse of the target object based on motion state information of the target object at a collision initial time, and expected motion state information of the target object at a collision end time; a collision point initial velocity determination module configured to determine an expected motion velocity of a collision point of the target object at the collision initial time based on the expected collision impulse of the target object, the expected motion state information of the target object at the collision end time, and expected motion state information of the robot at the collision end time; a robot initial velocity determination module configured to determine target motion state information of the robot at the collision initial time based on the expected motion velocity of the collision point of the target object at the collision initial time, and the motion state information of the target object at the collision initial time.

14. The robotic motion control system of claim 10, wherein, the collision post-motion type determination module comprises: a collision end motion state acquisition module configured to acquire motion state information of the collision point of the target object at the collision end time, and motion state information of the robot at the collision end time; a motion state comparison module configured to compare the motion state information of the collision point of the target object at the collision end time with the motion state information of the robot at the collision end time; a motion type determination module configured to determine a collision post-motion type of the target object based on a comparison result.

15. A robot, the robot comprising a robot, and wherein, the robot comprises a robot motion control system as claimed in any one of the preceding claims 10-14, and the control of the robot is implemented by a robot motion control method as claimed in any one of the claims 1-9.

Images