Robot returning method and system for space large-scale structure assembly

Abstract

The invention discloses a robot returning method for space large structure assembly, and relates to the technical field of space robots. The problems that an existing crawling type assembly robot is large in occupied space when crawling and returning, prone to collision and blockage and complex in task planning are solved. According to the technical scheme, after the robot completes assembly, a mechanical arm bounces to break away from the surface of the large space structure and obtains a designated initial speed pointing to a base station; then entering a free floating stage, firstly inhibiting the angular velocity through a controlled section and adjusting to a preset flight configuration, and then performing standby inertial flight in an uncontrolled section; a smooth track is generated through kinematics planning of discretization nonlinear programming in the whole process, it is ensured that base station capturing conditions are met in combination with sectional feed-forward PD control, and finally the track is captured by a base station.

Description

TECHNICAL FIELD

[0001] The application provides a robot returning method and system for assembling a large space structure, and belongs to the technical field of spaceflight. BACKGROUND

[0002] With the continuous progress of space technology and the gradual expansion of space applications, there is an increasing demand for large space structures such as space telescopes and large space stations. These large space structures are often large in mass and volume, and the current rocket launch technology cannot launch them from the ground to the designated working orbit at one time. Therefore, it is necessary to launch part of the components multiple times and assemble them in space, which puts forward higher requirements for the assembling technology of large space structures. The assembling technology of large space structures can automatically assemble various large structures in space, breaking the limitation that the required fuel increases exponentially with the increase of the launch mass of the rocket launch technology. It can save fuel costs while avoiding the risks brought by astronaut extravehicular assembly.

[0003] Currently, the assembling technology of large space structures mainly uses a crawling assembly robot. The crawling assembly robot is a robot that can move and assemble on the assembly structure. It has high reliability and assembly flexibility. The returning method of the existing crawling assembly robot still relies on the traditional crawling mechanism. During the returning process, the robot needs to continuously occupy the working space near the assembly structure. When multiple robots perform assembly operations in parallel, not only does it increase the complexity of task planning, but it also easily causes collisions and blockages between the robots due to their opposite movements. In addition, the crawling returning mode also has the problems of high energy consumption and low returning efficiency, which makes it difficult to adapt to the actual needs of efficient assembly of large space structures. SUMMARY

[0004] The application discloses a robot returning method for assembling a large space structure, which aims to solve the technical problems of the existing crawling assembly robot, such as large space occupation, easy collision and blockage with other robots, and complex task planning.

[0005] To solve the above technical problems, the technical solution adopted by the application is as follows: In a first aspect, the application provides a robot returning method for assembling a large space structure, comprising: S1, after the robot completes the assembly operation, the robot is separated from the surface of the large space structure by bouncing through the mechanical arm. When bouncing, one foot of the robot is fixed to the structure surface, each mechanical arm is quickly stretched to drive the center of mass to a specified initial speed pointing to the center of the base station, and then the fixed point is separated; S2, the robot enters a free floating stage, first quickly suppresses the joint angular velocity and adjusts to a preset flight configuration through a controlled section, and then enters an uncontrolled section to perform inertial flight in standby state; S3, during the whole bouncing and flying process, the robot state and input are discretized and nonlinearly planned by a kinematics planning module, a smooth motion trajectory is generated by combining polynomial interpolation, and a feedforward PD control module is used to output control signals in different stages to ensure that the robot inertia flying trajectory enters the base station capture space, the speed is within the base station capture range and there is no collision, until it is captured by the base station.

[0006] As a further improvement of the present application, the S1 specifically comprises: In the take-off stage, one foot of the robot is fixed on the rod, and the other foot can move freely in space; when taking off, the arms of the robot are quickly stretched upwards to accelerate the center of mass to the target speed, and then the foot on the rod is contacted and separated from the rod to enter the free floating stage.

[0007] As a further improvement of the present application, the conversion condition of the take-off stage in S1 and the free floating stage in S2 is that the center of mass acceleration of the robot perpendicular to the surface of the large structure in space is less than or equal to 0.

[0008] As a further improvement of the present application, the base station capture space in S3 is a sphere with a preset radius, the base station capture velocity set is a preset velocity range, and the robot inertia transfer initial velocity is within the capture velocity set and the trajectory completely falls into the capture space.

[0009] As a further improvement of the present application, the kinematics planning of the kinematics planning module is specifically: The state and input of the robot from the initial take-off to the end of the controlled segment are evenly discretized into N points according to time, the time step of adjacent points is set to a preset value, the actual motion state and input signal of the robot are generated by interpolating the discrete state points and input points by a polynomial of order S; The constraint conditions, stage conversion conditions and penalty functions of kinematics planning are determined to obtain an optimization objective of minimizing joint motor energy consumption and construct a target function; Wherein, the kinematics planning constraints include: dynamics constraints, velocity constraints, boundary constraints, mechanical arm collision constraints and time constraints.

[0010] As a further improvement of the present application, the kinematics planning of the kinematics planning module is specifically: The robot state and input are discretized into N points evenly divided according to time, the initial time is the first point, and the end of the controlled segment is the Nth point, and these points are used as variables for optimization; the time step , the discretized state and input are defined as follows:

[0011] In the formula, is time state variable, is time base pose quaternion, is time system input vector, is time joint angle vector. Polynomial is used to interpolate these discrete state points and input points to generate the actual state and input of robot

[0012] where s is the polynomial order, is the actual state of robot, is the actual input, and is the interpolation coefficient. The dynamics constraints include the take-off phase and the free-floating phase. Take-off phase: the dynamics constraint is

[0013] Free-floating phase: the dynamics constraint is

[0014] Velocity constraint: the error between the center of mass velocity and the target velocity should be less than a certain value, which is determined by the base manipulator workspace and the flight distance of the robot

[0015] The velocity of the robot is within a certain range

[0016] The robot should stop rotating at the end of the controlled segment, so

[0017] Boundary constraint: limited by the joint motor performance and the robot structure design, the joint torque, angle and angular velocity should meet the boundary conditions:

[0018] Manipulator collision constraint: no collision can be converted into the distance between the center lines of any two cylinders is not less than the sum of the radii of the two cylinders; Let the center line of the i-th envelope cylinder at time k be , and the cylinder radius be , then

[0019] where is the distance between two line segments.

[0020] Time constraint: set flight time , relationship

[0021] Phase transition condition: when the acceleration of the center of mass perpendicular to the surface direction of the truss is less than or equal to 0, the phase is converted, then

[0022] Penalty function:

[0023] The objective function is constructed as:

[0024] In the formula, N is the total number of discrete time steps, is the penalty function; is the joint torque; min is the minimum value, and max is the maximum value.

[0025] As a further improvement of the present application, the specific way of the feedforward PD control is: adopting independent joint control mode, the control frequency is 1kHz; the feedforward control in the take-off stage includes the center of mass Cartesian space PD control and the reference torque given by kinematics planning, the feedforward control frequency is 100Hz; The feedforward control of the free floating stage only contains the reference torque given by kinematics planning, and the goal of the angular velocity suppression of the controlled segment is to make the angular velocity of each joint of the robot approach to 0, and ensure the stability of the flight configuration.

[0026] As a further improvement of the present application, the feedforward PD control module is adopted to output control signals in different stages, including: Adopting independent joint PD control, the control frequency is 1kHz, and the control law is

[0027] The feedforward control in the take-off stage adds the center of mass Cartesian space PD control and the reference torque given by planning, the control frequency is 100Hz, and the control law is

[0028] Wherein the center of mass Jacobian matrix is:

[0029] In the free floating stage, the center of mass cannot be controlled, and the feedforward reference torque is

[0030] In the formula, is the PD control term of the control torque, and is a control coefficient matrix thereof, is a reference joint angle, is a feedforward control item of a control torque, and is a control coefficient matrix thereof, is a center-of-mass reference displacement, is a center-of-mass actual displacement, is a reference torque, 、 、 and are a center-of-mass displacement of the ith link, an upper joint axis unit vector, an upper joint center displacement, and a link mass, respectively, is a total mass.

[0031] As a further improvement of the present application, the preset flight configuration is a symmetrical configuration facilitating recognition by the base station through the mechanical arm capturing the robot in inertia flight in the uncontrolled section within the working space of the base station.

[0032] In a second aspect, the present application provides a robot returning system for assembling a large space structure, comprising an assembling robot body, a base station, and a mechanical arm for driving the robot to move, comprising: The robot returns in an inertia transfer returning mode, and the mechanical arm is configured to make the robot leave the surface of the large space structure through bouncing and obtain a specified initial velocity pointing to the center of the base station. The bouncing process is divided into a take-off stage and a free floating stage. In the take-off stage, one foot of the robot is fixed to the surface of the structure, and after each arm is stretched to drive the center of mass to accelerate to a target velocity, the robot is separated from the fixed point. The free floating stage includes a controlled section connected to the take-off stage and a subsequent uncontrolled section. The controlled section makes the robot enter a specified flight configuration through angular velocity suppression, and the robot in the uncontrolled section is in a standby state for inertia flight until it is captured by the base station. The robot is configured with a kinematics planning module and a feedforward PD control module. The kinematics planning module discretizes the robot state and input into N time points and generates a smooth motion trajectory through polynomial interpolation, and incorporates dynamics constraints, speed constraints, boundary constraints, mechanical arm collision constraints, and time constraints. The feedforward PD control module differentiates the feedforward strategy according to the take-off stage and the free floating stage, independently controls each joint, and ensures that the inertia flight trajectory of the robot enters the captureable space of the base station, the speed is within the captureable range of the base station, and there is no collision.

[0033] The present application has the following beneficial effects relative to the prior art: The traditional crawling return is changed to inertial transfer return, the robot standby inertial flight does not need to occupy the assembly area space, fully utilizes the idle space near the large structure, reduces the collision and congestion risk of multi-robot parallel operation from the root, simplifies the task planning process; the specified initial speed is obtained through the mechanical arm bounce, the energy consumption and time cost of the return process are greatly reduced, and the return efficiency is improved; the combination of kinematics planning and segmented feedforward PD control ensures that the robot meets the trajectory, speed and collision-free requirements of the base station capture, and guarantees the stability and reliability of the return process; adapt to the complex operation environment of large space structure assembly, provide technical support for efficient and safe assembly of large space structure. BRIEF DESCRIPTION OF DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following introduces the drawings of the related technical solutions in the embodiments of the present application or the prior art. It should be understood that the drawings in the following introduction are only for the convenience of clearly expressing part of the embodiments of the technical solutions of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the premise of the drawings.

[0035] Figure 1 The robot return control principle diagram for large space structure assembly; Figure 2 The robot kinematics planning curve diagram; wherein, (a) is the center of mass velocity curve diagram; (b) is the joint angle curve diagram; (c) is the joint angular velocity curve diagram; (d) is the joint torque curve diagram; Figure 3 The robot planning result control tracking curve diagram; wherein, (a) is the joint torque curve diagram; (b) is the center of mass velocity curve diagram; (c) is the joint angle curve diagram; (d) is the joint angular velocity curve diagram. DETAILED DESCRIPTION

[0036] The embodiments of the present application are described in detail below, examples of which are shown in the drawings, wherein the same or similar reference signs represent the same or similar elements or elements with the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary, only for explaining the present application, and cannot be understood as limiting the present application. For the step numbers in the following embodiments, only for the convenience of explanation, the order between the steps is not limited, and the execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

[0037] In the description of the present application, unless otherwise explicitly limited, the words such as setting, installing, connecting, etc. should be understood in a broad sense, and those skilled in the art can reasonably determine the specific meaning of the above words in the present application in combination with the specific content of the technical solutions.

[0038] The first purpose, such as Figure 1 As shown, this application provides a robot return device for assembling large-scale spatial structures, including an assembly robot body, a base station, and a robotic arm for driving the robot's movement. The robot uses an inertial transfer return method instead of crawling return. The robotic arm is configured to allow the robot to detach from the surface of the large-scale spatial structure and obtain a specified initial velocity pointing towards the center of the base station via a jump. The jump process is divided into a take-off phase and a free-floating phase. In the take-off phase, one foot of the robot is fixed to the structural surface, and each arm extends to drive the center of mass to accelerate to the target velocity before detaching from the fixed point. The free-floating phase includes a controlled segment connecting the take-off phase and a subsequent uncontrolled segment. In the controlled segment, the robot enters a specified flight configuration by suppressing angular velocity. In the uncontrolled segment, the robot is in a standby state and performs inertial flight until it is captured by the base station. The robot is equipped with a kinematics planning module and a feedforward PD control module. The kinematics planning module discretizes the robot's state and input into N time points and generates a smooth motion trajectory through polynomial interpolation, incorporating dynamics, velocity, boundary, collision, time constraints, and stage transition conditions. The feedforward PD control module sets a feedforward strategy differently for the take-off phase and the free-floating phase, at 1kHz. The frequency is used to independently control each joint, ensuring that the robot's inertial flight trajectory enters the space that the base station can capture, the speed is within the range that the base station can capture, and there is no collision.

[0039] This application transforms the traditional crawling return into an inertial transfer return. During the inertial transfer return, the robot remains in standby flight without engaging in any activity, reducing the time and energy consumption of slow crawling return. It also makes full use of unoccupied space near large structures, avoids space occupation issues caused by the return, reduces the task planning complexity caused by a large number of robots performing assembly work in parallel, and reduces the risk of robot collision and blockage events.

[0040] Secondly, this application provides a robot return method for assembling large-scale space structures, comprising the following steps: S1. After the robot completes the assembly work, it uses the robotic arms to jump and detach from the surface of the large spatial structure. During the jump, one of the robot's feet is fixed to the surface of the structure, and each robotic arm quickly extends to drive the center of mass to accelerate to the specified initial velocity pointing to the center of the base station, and then detaches from the fixed point. The transition condition between the jump phase in S1 and the free-floating phase in S2 is that the acceleration of the robot's center of mass perpendicular to the surface of the large spatial structure is less than or equal to 0.

[0041] In S3, the space that the base station can capture is a sphere with a preset radius, the set of speeds that the base station can capture is a preset speed range, and the robot's initial velocity of inertial transfer is within the set of speeds that can be captured and the trajectory falls completely into the space that can be captured.

[0042] S2, the robot enters a free floating stage, first quickly suppresses the angular velocity of each joint through the controlled section and adjusts to the preset flight configuration, and then enters the uncontrolled section to perform inertial flight in standby state; S3, during the entire bouncing and flying process, the robot state and input are discretized and nonlinearly planned through the kinematic planning module, a smooth motion trajectory is generated by combining polynomial interpolation, and a feedforward PD control module is used to output control signals in different stages to ensure that the inertial flight trajectory of the robot enters the capture space of the base station, the speed is within the capture range of the base station, and there is no collision, until the robot is captured by the base station.

[0043] Specifically, as an optional solution, the specific process of kinematic planning of the kinematic planning module is to discretize the state and input of the robot from the initial take-off to the end of the free floating controlled section into N points uniformly in time, set the time step of adjacent points to a preset value, and interpolate the discrete state points and input points by a polynomial of order S to generate the actual motion state of the robot and the input signal.

[0044] The velocity constraint of the kinematic planning satisfies that the error between the center of mass velocity of the robot and the target velocity is less than a preset threshold, which is determined by the base station mechanical arm workspace and the flight distance of the robot; the robot motion velocity is adapted to the sensor accuracy and the mechanical arm control response speed; the angular velocity of the robot at the end of the free floating controlled section is 0.

[0045] The collision constraint of the kinematic planning is realized by the following way: setting a minimum envelope cylinder for each part of the robot mechanical arm, the distance between the center lines of any two envelope cylinders is not less than the sum of the radii of the two cylinders, and the center line is the connection line of the centers of the cylinder end faces.

[0046] The detection object of the collision constraint includes the truss of a large space structure, and the envelope cylinder is set for the truss to avoid collision between the robot in the free floating stage and the truss.

[0047] Specifically, as an optional solution, the specific way of the feedforward PD control is: using independent joint control mode, the control frequency is 1 kHz; the feedforward control in the take-off stage includes the center of mass Cartesian space PD control and the reference torque given by the kinematic planning, and the feedforward control frequency is 100 Hz.

[0048] The feedforward control of the free floating stage only includes the reference torque given by the kinematic planning, and the target of the angular velocity suppression in the controlled section is to make the angular velocity of each joint of the robot approach to 0 to ensure the stability of the flight configuration.

[0049] The preset flight configuration is a symmetrical configuration convenient for the base station mechanical arm to identify, and the base station captures the robot in inertial flight in the uncontrolled section through the mechanical arm in its workspace.

[0050] The specific solutions of the present application are described below in combination with specific embodiments: The robot inertia transfer return needs to reach a specified initial speed. The relative motion relationship between the robot and the mass center after the robot leaves the surface of the large space structure can be obtained by the C-W equation:

[0051] In the formula, and are the position of the robot relative to the mass center of the large space structure and the orbital angular velocity of the large space structure, respectively. Since the inertia transfer return time is short and the orbital angular velocity of the large space structure is small, the C-W equation can be approximately degenerated into the equation of uniform straight line motion:

[0052] It can be seen that each standby flight trajectory is uniquely determined by the structure leaving position and the inertia transfer return initial speed, and the generation function is denoted as , and are the vector and the final speed of the mass center of the robot relative to the mass center of the large space structure during the inertia transfer return.

[0053] The inertia transfer return process needs to ensure that the inertia transfer trajectory of the robot after leaving the structure surface meets the base station capture condition: 1. The inertia transfer trajectory enters the base station captureable space; 2. The inertia transfer initial speed is within the base station captureable speed range; 3. The inertia transfer trajectory is within the non-obstructed trajectory set of the structure leaving region.

[0054] Assuming that the captureable space is a sphere with a radius , the captureable speed set , and the non-obstructed trajectory set of the structure leaving region is , the capture condition can be expressed as:

[0055] To ensure that the robot meets the capture condition and leaves a margin to ensure that it can still be received by the base station after being disturbed by factors such as perturbation, the inertia transfer initial speed needs to be directed to the center of the base station.

[0056] The way the robot leaves the surface of the structure and reaches the specified initial speed adopts the mechanical arm bouncing method. The robot bouncing process is divided into two stages, namely the take-off stage and the free floating stage.

[0057] In the take-off stage, one foot of the robot is fixed on the rod, and the other foot can move freely in space.

[0058] At the moment of take-off, the robot arms stretch upward rapidly to accelerate the center of mass to the target velocity, and then the feet on the bar contact the fixed bar to enter the free-floating stage.

[0059] During the whole take-off stage, the feet are fixed on the bar, so the ZMP problem does not need to be considered. The free-floating stage is divided into a controlled segment and an uncontrolled segment. The controlled segment connects the take-off stage, and the robot joints still have angular velocity, so the controlled segment should quickly suppress the angular velocity and enter the specified flight configuration to suppress its free rolling in space, which is convenient for the base station mechanical arm to identify and capture. After entering the specified flight configuration, i.e., the uncontrolled segment, the robot will enter the standby state for inertial flight until it is captured by the base station.

[0060] To make the robot center of mass motion velocity reach the specified initial velocity, the robot manipulator should be kinematically planned and controlled. Kinematic planning is a continuous trajectory optimization problem, which can be converted into a nonlinear programming problem by discretization for solution. The planning starts from the initial take-off and ends at the end of the free-floating controlled segment. The two stages are jointly planned, and the phase is converted through specific conditions during the planning process.

[0061] Variable definition: The robot state and input are discretized into N points evenly divided by time, with the initial time as the first point and the end of the controlled segment as the Nth point. These points are used as variables for optimization. The time step is defined as

[0062] In the formula, is the state variable at time , is the base attitude quaternion at time , and is the joint angle at time

[0063] To ensure the smoothness of the actual state variables and input variables, polynomials are used to interpolate these discrete state points and input points to generate the actual state of the robot and the input:

[0064] where s is the polynomial order.

[0065] The constraints of kinematic planning include dynamics constraints, velocity constraints, boundary constraints, manipulator collision constraints, and time constraints. The following will be described respectively: The dynamics constraints include the take-off stage and the free-floating stage.

[0066] ​​​Take-off stage: Since the robot has one foot fixed in the take-off stage, the fixed foot can be regarded as a fixed base, so the robot is a fixed-base robot, and the end is not forced, and the relationship between joint angular acceleration and joint torque can be obtained from the robot dynamics equation

[0067] The above formula can be written as

[0068] Therefore, there is a dynamic constraint

[0069] Free floating stage: The robot dynamics equation in the free floating stage is as follows

[0070] The dynamics equation shows that there is also a dynamic constraint ; Velocity constraint: The robot bounces towards the base station, and the relative motion relationship with the truss structure can be given by the C-W equation

[0071] When the orbital angular velocity of the truss structure is very small, the equation degenerates into the equation of uniform linear motion

[0072] The base station manipulator has a certain working space, and the robot can only be captured when it enters its working space, so the error between the center of mass velocity and the target velocity should be less than a certain value, which is determined by the working space of the base station manipulator and the flight distance of the robot

[0073] Due to the accuracy of the sensor and the control response speed of the manipulator, the velocity of the robot should be within a certain range

[0074] The robot should stop rotating at the end of the controlled section, that is

[0075] Boundary constraint: Due to the joint motor performance and the design of the robot structure, the joint torque, angle and angular velocity should meet the boundary conditions

[0076] Mechanical arm collision constraint: due to the high degree of freedom of the mechanical arm, the mechanical arm itself may collide during the rebound, causing damage, so the movement of the mechanical arm should be collision-free. Taking the minimum envelope cylinder of each part of the mechanical arm, the collision-free can be converted into the distance between the center lines (the center of the end face of the cylinder) of any two cylinders is not less than the sum of the radii of the two cylinders. Let the center line of the i-th envelope cylinder at time k be , and the radius of the cylinder be , then

[0077] , where is the distance between two line segments.

[0078] At the same time, the robot cannot collide with the truss after entering the free floating stage, otherwise the collision will change the linear momentum and angular momentum, and the robot only has internal forces in the free floating stage, which cannot be corrected, so the truss envelope cylinder is also included in the above formula.

[0079] Time constraint: the robot should stop rotating before being recognized by the mechanical arm, so the duration of the controlled segment cannot be longer than the flight time. To leave enough margin, the entire process time from the start of the take-off to the end of the controlled segment is shorter than the flight time, and the flight time is , and the relationship is

[0080] Phase transition condition: when the acceleration of the center of mass perpendicular to the surface of the truss is less than or equal to 0, the phase is converted, i.e.

[0081] Penalty function: to reduce the energy consumption of the joint motor, the energy consumption penalty function is set as

[0082] In summary, the mathematical description of the NLP problem is

[0083] where N is the total number of discrete time steps, is the penalty function; is the joint torque; min is the minimum value, and max is the maximum value.

[0084] The robot control uses feedforward PD control, where the feedforward calculation method is different in the take-off stage and the free floating stage due to the change of the control scenario.

[0085] Independent PD control is adopted for each joint, with a control frequency of 1 kHz, and the control law is

[0086] Feedforward: The feedforward adds the reference torque given by the centroid Cartesian space PD control and planning to the take-off phase control, the control frequency is 100Hz, and the control law is:

[0087] In the formula, is the PD control term of the control torque, and is the parameter matrix of the PD control, is the reference joint angle, is the feedforward control term of the control torque, and is the parameter matrix of the feedforward control, is the centroid reference displacement, is the centroid actual displacement, is the reference torque.

[0088] Where the centroid Jacobian matrix is:

[0089] In the formula, , , is the mass, centroid displacement and upper joint axis unit vector of the ith connecting rod, is the upper joint center displacement.

[0090] In the free floating phase, since the centroid cannot be controlled, the feedforward only has the reference torque:

[0091] The application also simulates the above scheme, and the simulation process is as follows: Suppose the robot is located on a rod, and a coordinate system is established at the position of the robot, the Z direction is perpendicular to the rod upward, the base station is located at , and there is no obstacle between the base station and the origin to block the planning time .

[0092] The robot structure characteristics are shown in the following table

[0093] According to the capture condition, the inertial transfer return initial velocity of the robot can be

[0094] The kinematic planning of the robot is obtained, and the results are shown in Figure 2 . The curve is relatively smooth and only has a mutation at the switching point, and the change of each joint angle is within the allowable range, and the centroid velocity error is:

[0095] The control tracking of the above planning result is shown in Figure 3 It can be seen that the overall tracking accuracy is high except for a small part of tracking error, and the mass center velocity error is only .

[0096] A third object of the embodiments of the present application is to provide an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-mentioned robot returning method for assembling a large structure in space. The electronic device further comprises a communication interface and a bus.

[0097] A fourth object of the embodiments of the present application is to provide a computer readable storage medium, which stores a computer program, wherein the computer program is executed by a processor to implement the above-mentioned robot returning method for assembling a large structure in space.

[0098] A fifth object of the embodiments of the present application is to provide a computer program product, which comprises computer instructions, wherein the computer instructions instruct a computer to execute the above-mentioned robot returning method for assembling a large structure in space.

[0099] These computer program instructions can also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce a manufactured product comprising instruction devices, which implement the functions specified in the flow Figure 1 one or more flows and / or blocks Figure 1 one or more blocks or multiple blocks.

[0100] These computer program instructions can also be loaded into a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to produce a computer implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in the flow Figure 1 one or more flows and / or blocks Figure 1 one or more blocks or multiple blocks.

[0101] The present application can adopt a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer usable storage media (including but not limited to disk storage, readable storage media, optical storage, etc.) containing computer usable program code.

[0102] The computer program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other processing device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other processing device to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions specified in the flowchart block or blocks. Figure 1 The flowchart and / or block diagram in the variation of the present application can describe any device, system or computer program products that perform the specified steps and / or functions as a means for obtaining the processes specified in the flowchart and / or block diagram. Figure 1 The device configured to perform the functions specified in the flowchart and / or block diagram of one or more steps and / or blocks.

[0103] Obviously, the embodiments described above are only a part rather than all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work should belong to the protection scope of the present application.

[0104] Finally, it should be noted that the above embodiments are used to illustrate the technical solutions of the present application rather than limit the technical solutions of the present application. Although the present application has been described in detail with reference to the above embodiments, a person of ordinary skill in the art should understand that the specific implementation manners of the present application can be modified or replaced equivalently without departing from the spirit and scope of the present application, and any modification or equivalent replacement should be covered in the protection scope of the claims of the present application.

Claims

1. A robot return method for assembling large-scale spatial structures, characterized in that, Includes the following steps: S1. After the robot completes the assembly work, it uses the robotic arms to jump and detach from the surface of the large spatial structure. During the jump, one of the robot's feet is fixed to the surface of the structure, and each robotic arm quickly extends to drive the center of mass to accelerate to the specified initial velocity pointing to the center of the base station, and then detaches from the fixed point. S2, the robot enters the free-floating stage. First, it quickly suppresses the angular velocity of each joint and adjusts to the preset flight configuration through the controlled segment, and then enters the uncontrolled segment to perform inertial flight in standby mode. S3, throughout the entire bouncing and flying process, uses the kinematic planning module to perform discretized nonlinear programming on the robot's state and input, and combines it with polynomial interpolation to generate a smooth motion trajectory. At the same time, the feedforward PD control module outputs control signals in stages to ensure that the robot's inertial flight trajectory enters the base station's capture space, its speed is within the base station's capture range, and there is no collision, until it is captured by the base station.

2. The robot return method for assembling large spatial structures according to claim 1, characterized in that: S1 specifically includes: During the take-off phase, one of the robot's legs is fixed to the pole, while the other leg can move freely in space. When taking off, the robot's arms quickly extend upwards, accelerating the center of mass to the target speed. After that, the leg on the pole contacts the fixed pole and detaches, entering the free-floating phase.

3. The robot return method for assembling large spatial structures according to claim 1, characterized in that: The transition condition between the jump phase in S1 and the free-floating phase in S2 is that the acceleration of the robot's center of mass perpendicular to the surface of the large spatial structure is less than or equal to 0.

4. The robot return method for assembling large spatial structures according to claim 1, characterized in that: In S3, the space that the base station can capture is a sphere with a preset radius, the set of speeds that the base station can capture is a preset speed range, and the robot's initial velocity of inertial transfer is within the set of speeds that can be captured and the trajectory falls completely into the space that can be captured.

5. The robot return method for assembling large spatial structures according to any one of claims 1 to 4, characterized in that: The specific process of kinematic planning in the kinematic planning module is as follows: The robot's state and input from the initial jump to the end of the controlled free-floating phase are uniformly discretized into N points over time. The time steps of adjacent points are set to preset values. The discrete state points and input points are interpolated by a polynomial of order S to generate the robot's actual motion state and input signal. The constraints, phase transition conditions, and penalty functions of kinematic programming are determined, and the objective function is constructed with the goal of minimizing the energy consumption of the joint motors. The constraints in kinematic programming include: dynamic constraints, velocity constraints, boundary constraints, robot arm collision constraints, and time constraints.

6. The robot return method for assembling large spatial structures according to claim 5, characterized in that: The specific process of kinematic planning in the kinematic planning module is as follows: The robot's state and input are discretized into N points evenly divided over time, with the initial time being the first point and the end time of the controlled segment being the Nth point. These points are then used as variables for optimization. A time step is defined. The discretized state and input are defined as follows: In the formula, for Time-state variables, for Timing base attitude quaternion, for The system input vector at time step, for At any given moment, the joint angle vector; Polynomial interpolation is used to generate the robot's actual state and input by interpolating these discrete state points and input points. In the formula, s is the order of the polynomial. This represents the robot's actual state. For actual input, and These are the interpolation coefficients; The dynamic constraints include the takeoff phase and the free-floating phase; Takeoff phase: Dynamic constraints are Free-floating phase: dynamic constraints are Velocity constraint: The error between the center-of-mass velocity and the target velocity should be less than a certain value, which is determined by the base station robotic arm's workspace and the robot's flight distance. The robot's speed is within a certain range The robot should stop rotating at the end of the controlled segment. Boundary constraints: Due to limitations imposed by the joint motor performance and robot structural design, the joint torque, angle, and angular velocity must all meet the boundary conditions. Robotic arm collision constraint: Preventing collisions can be transformed into ensuring the distance between the centerlines of any two cylinders is not less than the sum of their radii; let the centerline of the i-th enclosing cylinder at time k be... Its cylinder radius is ,but in The distance between the two line segments; Time constraint: Set flight time There is a relationship Stage transition condition: The stage transition occurs when the acceleration of the center of mass perpendicular to the truss surface is less than or equal to 0. Penalty function: The objective function is constructed as follows: In the formula, N It is the total number of discrete time steps. For the penalty function; Joint torque; min is the minimum value, and max is the maximum value.

7. The robot return method for assembling large spatial structures according to claim 1, characterized in that: The specific method of the feedforward PD control is as follows: each joint adopts an independent control mode with a control frequency of 1kHz; the feedforward control during the take-off phase includes centroid Cartesian space PD control and reference torque given by kinematic planning, with a feedforward control frequency of 100Hz. The feedforward control in the free-floating phase only includes the reference torque given by the kinematics planning, and the goal of the controlled segment angular velocity suppression is to make the angular velocities of each joint of the robot approach 0, so as to ensure the stability of the flight configuration.

8. The robot return method for assembling large-scale space structures according to claim 1, characterized in that: The method of using a feedforward PD control module to output control signals in stages with differentiated outputs includes: Independent PD control is used for each joint, with a control frequency of 1kHz and a control law of... During the takeoff phase, the control feedforward incorporates the reference torque provided by the centroid Cartesian space PD control and planning, with a control frequency of 100Hz and the control law being... Where the centroid Jacobian matrix is: During the free-floating phase, since the center of mass cannot be controlled, the feedforward reference torque is... In the formula, For the PD control term that controls torque, and It is its control coefficient matrix. It refers to the joint angle. It is a feedforward control term for controlling torque. and It is its control coefficient matrix. It is the centroid reference displacement. It is the actual displacement of the center of mass. It is a reference torque. , , and These are, respectively, the displacement of the center of mass of the i-th link, the unit vector along the axis of the upper joint, the displacement of the center of the upper joint, and the mass of the link. It refers to the total mass.

9. The robot return method for assembling large spatial structures according to claim 1, characterized in that: The preset flight configuration is a symmetrical configuration that is easy for the base station's robotic arm to recognize. The base station captures the robot in uncontrolled inertial flight within its workspace using the robotic arm.

10. A robot return system for assembling large-scale spatial structures, comprising an assembly robot body, a base station, and a robotic arm for driving the robot's movement, characterized in that, include: The robot returns via inertial transfer. The robotic arm is configured to bounce the robot off the surface of a large spatial structure and gain a specified initial velocity pointing towards the center of the base station. The bouncing process is divided into a take-off phase and a free-floating phase. In the take-off phase, one foot of the robot is fixed to the structural surface, and the arms extend to accelerate the center of mass to the target velocity before leaving the fixed point. The free-floating phase includes a controlled segment connecting the take-off phase and a subsequent uncontrolled segment. In the controlled segment, the robot enters a specified flight configuration by suppressing angular velocity. In the uncontrolled segment, the robot is in a standby state and performs inertial flight until it is captured by the base station. The robot is equipped with a kinematics planning module and a feedforward PD control module. The kinematics planning module discretizes the robot's state and input into N time points and generates a smooth motion trajectory through polynomial interpolation, incorporating dynamic constraints, velocity constraints, boundary constraints, robotic arm collision constraints, and time constraints. The feedforward PD control module sets different feedforward strategies for the take-off phase and the free-floating phase, and independently controls each joint to ensure that the robot's inertial flight trajectory enters the base station's capture space, its velocity is within the base station's capture range, and there is no collision.

Images