A decentralized collaborative transport method for a dragger-assistant system

Abstract

This invention belongs to the field of robotics and discloses a decentralized collaborative handling method for a dragger-assistant system. The method includes: first, designing a decentralized estimator for the assistor to estimate the desired trajectory and dynamic parameters; then, developing a decentralized controller to achieve collaborative handling; and finally, employing a dynamic regression extension and hybrid DREM method to estimate unknown parameters, ensuring that the parameters converge to their true values ​​under continuous excitation conditions. This invention proposes a decentralized collaborative handling method for a dragger-assistant system, dividing the roles of the robots into draggers and assistors and integrating the DREM parameter estimation method to solve the trajectory tracking problem.

Description

Technical Field

[0001] The present invention belongs to, but is not limited to, the field of robotics, and in particular relates to a decentralized collaborative handling method and system for a dragger-assistant system. Background Art

[0002] Robots are a promising platform for replacing humans in heavy tasks such as carrying and lifting. Similar to how multiple people collaborate to complete a task, collaborative lifting can be accomplished using a team of robots, which requires research into coordination mechanisms and controller design.

[0003] Collaborative handling has been extensively studied in the literature, with two main approaches emerging. The first approach, similar to the multi-agent problem, enables collaborative handling through communication between robots, where communication delays and failures must be considered for system robustness. The second approach involves a class of decentralized control algorithms that do not require a communication network; however, a global planner is required to send the object's motion plan to each robot.

[0004] Two issues complicate collaborative handling tasks. First, the object's dynamic parameters are often unknown. This problem can be addressed using traditional adaptive control laws, such as model reference adaptive control, but these methods are inapplicable when each robot lacks knowledge of its trajectory in the absence of a centralized planner. Second, posture control is crucial for collaborative handling trajectory tracking but has yet to be fully explored. Summary of the Invention

[0005] In view of the problems existing in the prior art, the present invention provides a decentralized collaborative transport method for a dragger-assistant system.

[0006] The present invention is implemented as follows: a decentralized collaborative transport method for a dragger-assistant system, the method comprising:

[0007] S1: First, a decentralized estimator is designed for the helper to estimate the desired trajectory and dynamic parameters, and then a decentralized controller is proposed to achieve cooperative handling;

[0008] S2: Dynamic regression extension and hybrid DREM method are used to estimate unknown parameters, ensuring that the parameters converge to the true value under the condition of continuous excitation.

[0009] Furthermore, the method specifically includes:

[0010] 1. Model the overall collaborative handling system and describe the problems and tasks.

[0011] Assume {I} is an inertial system, {B} is a fixed system attached to the center of mass CoM of the rigid body; a group of N robots are in a rigid contact state with the rigid body; since it is usually difficult to obtain the exact position of the rigid body CoM, a fixed system located at {B}r is introduced. p The auxiliary frame {P} is aligned with {B}. The origin of {P} is used as the anchor point. In real-world applications, a marker can be placed at the anchor point, which can be observed by each robot.

[0012] For the i-th robot, r i Defined as the known vector from the anchor point to the contact position of the ith robot; let x∈R3,R∈SO(3) be the position and posture of {P} respectively; the dynamics of the collaborative handling system is described as

[0013]

[0014] where q = [x, R] ∈ R 3 ×SO(3), and are the generalized configuration variables, velocity and acceleration of the anchor point, is the force and torque applied by the ith robot to the rigid body; in formula (1), is the inertia matrix, For centrifugal and Coriolis matrices, is the gravity vector; for simplicity, H, C, and g are sometimes used instead 、 and ; Assume that all robots are unaware of the H, C, and g matrices.

[0015]

[0016] is the so-called grasping matrix, which transforms the force / torque exerted by the ith robot at the contact position into an effect denoted by {P};

[0017] The Euler-Lagrangian system (1) has the following properties:

[0018] P1: Inertia matrix is positive definite and bounded.

[0019] P2: Centrifugal Coriolis Matrix It is bounded.

[0020] P3: Matrix-valued functions is obliquely symmetric, that is:

[0021]

[0022] P4: There exists a constant vector and a matrix function , making

[0023]

[0024] in , is called the regressor;

[0025] From the dynamics (1) and expression (6) in P4, we get

[0026]

[0027] in , represents the unknown parameters in the system dynamics equations.

[0028] This invention is oriented to trajectory tracking tasks and designs a decentralized controller for each robot. , without relying on robot-to-robot communication, to carry rigid objects and track the desired trajectory. The desired trajectory is described as

[0029]

[0030] in are the desired position, acceleration, angular velocity and attitude respectively; let , The system and the desired trajectory satisfy:

[0031] Assumption 1: The collaborative handling system - (1) satisfies properties P1-P4;

[0032] Assumption 2: The desired trajectory in is only for the first robot Available; Status and Accessible to all robots.

[0033] 2. Design of Expected Trajectory Estimator and Parameter Estimator

[0034] Furthermore, the estimation of the desired trajectory and system parameters specifically includes:

[0035] The expected trajectory estimator is designed for each helper as follows:

[0036]

[0037]

[0038] in , , and They are Estimates of the desired position, velocity, attitude, and angular velocity of each robot, and is the positive estimated gain; for simplicity, let ;

[0039] The wavy lines on the letters represent the error of the corresponding variable relative to the true quantity, and the corresponding error dynamics equation is obtained:

[0040]

[0041]

[0042] Where A is the Hurvez matrix, s x are the components of the following composite tracking error s.

[0043]

[0044] in It is called the reference speed and is defined as

[0045]

[0046] In the formula represents the posture tracking error;

[0047] Next, the DREM method is used to design parameter estimators for each robot, specifically and As shown below

[0048]

[0049] in is the estimated parameter of the ith robot, is a positive definite matrix, Given in P4, is a positive scalar; is the DREM regressor, which can be derived through the DREM algorithm.

[0050] 3. Controller Design

[0051] Define the estimated reference velocity and the estimated composite error as

[0052]

[0053] Then, the decentralized controllers for all robots are designed as follows:

[0054]

[0055]

[0056]

[0057] in is the formula (2) The inverse matrix of, for i∈V and h,k i is a positive scalar gain; note that in (35)-(36) we use and , to represent and . Define the tracking error and obtain the error dynamics as

[0058]

[0059]

[0060] .

[0061] 4. Proof of Stability

[0062] The Lyapunov function is designed as follows: ,in , is the Lyapunov equation The solution of is a symmetric positive definite matrix. , , , It can be shown that when other gains are given, the gains and and When the value is large enough, trajectory tracking is achieved. In addition, if the regressor When the excitation is continuous, the parameter estimation error converges to zero.

[0063] Another object of the present invention is to provide a decentralized collaborative transport system for a dragger and assistor system based on the decentralized collaborative transport method for the dragger and assistor system, the system specifically comprising:

[0064] The desired trajectory estimator module first designs a decentralized estimator for the assistant to estimate the desired trajectory in order to implement control based on it;

[0065] The parameter estimation module uses dynamic regression expansion and hybrid DREM method to estimate unknown parameters, ensuring that the parameters converge to the true value under the condition of continuous excitation;

[0066] The controller module is connected with the desired trajectory estimator module and the parameter estimation module. Based on the desired trajectory estimator and the parameter estimator, a distributed controller of the robot is designed to realize collaborative handling.

[0067] An embodiment of the present invention provides a computer device, which includes a memory and a processor. The memory stores a computer program. When the computer program is executed by the processor, the processor executes the steps of the decentralized collaborative transport method of the dragger-assistant system.

[0068] An embodiment of the present invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the steps of the decentralized collaborative transport method of the dragger-assistant system.

[0069] An embodiment of the present invention provides an information data processing terminal, which is used to implement the decentralized collaborative transport system of the dragger-assistant system.

[0070] In combination with the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solutions to be protected by the present invention are as follows:

[0071] First, the present invention proposes a decentralized collaborative handling method for a dragger-helper system, dividing the roles of robots into draggers and helpers and integrating the DREM parameter estimation method to solve the trajectory tracking problem. This invention can solve the problem of requiring presetting reference trajectories for all robots in existing collaborative handling solutions, thereby avoiding 1) the redundant work of timestamp alignment for distributed multi-robot groups without communication in real-world applications; and 2) the resetting of reference trajectories for the entire multi-robot group for different tasks. This can significantly reduce the operational difficulty of decentralized multi-robot groups and improve task adaptability. The present invention solves the aforementioned problems by designing a decentralized expected trajectory estimator and incorporating it into the control system stability analysis, ultimately resulting in a decentralized multi-robot collaborative handling framework that theoretically meets stability requirements and reduces the application threshold in reality.

[0072] Second, as auxiliary evidence of the invention's inventiveness, it is also reflected in the following important aspects:

[0073] (1) The expected benefits and commercial value of the technical solution of the present invention after transformation are:

[0074] This invention provides a collaborative handling solution for factory applications, which can greatly lower the threshold for multiple robots to collaboratively carry heavy objects in factory applications, provide a solution for the realization of digital and intelligent unmanned (or low-manned) factories, and reduce labor costs.

[0075] (2) The technical solution of the present invention solves a technical problem that people have long been eager to solve but have never been able to solve successfully:

[0076] Unmanned factory. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 This is a flow chart of a distributed collaborative transport method of a dragger-assistant system provided by an embodiment of the present invention;

[0078] Figure 2 This is a schematic diagram of the collaborative robot handling principle provided by an embodiment of the present invention;

[0079] Figure 3 is a schematic diagram of the proposed framework provided by an embodiment of the present invention;

[0080] Figure 4 The simulation results of six robots collaboratively carrying a rigid object provided by the embodiment of the present invention are as follows: (a) s norm; (b) x e norm; (c) tr(I3−R e ); (d)q˙ e The norm of

[0081] Figure 5 is a simulation result of parameter estimation of the DREM method provided by an embodiment of the present invention;

[0082] Figure 6 It is the physical model in Gazebo provided by the embodiment of the present invention

[0083] Figure 7 1 and 2 are physical simulation results provided by an embodiment of the present invention: (a) position; (b) velocity; (c) angle; and (d) angular velocity. DETAILED DESCRIPTION

[0084] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0085] like Figure 1 As shown, an embodiment of the present invention provides a decentralized collaborative transport method of a dragger-assistant system, the method comprising:

[0086] S1: First, a decentralized estimator is designed for the helper to estimate the desired trajectory and dynamic parameters, and then a decentralized controller is proposed to achieve cooperative handling;

[0087] S2: Dynamic regression extension and hybrid DREM method are used to estimate unknown parameters, ensuring that the parameters converge to the true value under the condition of continuous excitation.

[0088] The method specifically includes:

[0089] 1. Model the overall collaborative handling system and describe the problems and tasks.

[0090] The schematic diagram of collaborative robot handling is as follows Figure 2 As shown in the figure, let {I} be the inertial system and {B} be the fixed system attached to the center of mass CoM of the rigid body. A group of N robots are in a rigid contact state with the rigid body. Since it is usually difficult to obtain the exact position of the rigid body CoM, an auxiliary system {P} is introduced at {B}rp, and its direction is aligned with {B}. The origin of {P} is used as the anchor point. In real-world applications, a marker can be placed at the anchor point, which can be observed by each robot.

[0091] For the ith robot, define ri as the known vector from the anchor point to the contact position of the ith robot; let x∈R3,R∈SO(3) be the position and posture of {P} respectively; the dynamics of the collaborative handling system is described as

[0092]

[0093] where q = [x, R] ∈ R3 × SO(3), and are the generalized configuration variables, velocity and acceleration of the anchor point, is the force and torque applied by the ith robot to the rigid body; in formula (1), is the inertia matrix, For centrifugal and Coriolis matrices, is the gravity vector; for simplicity, H, C, and g are sometimes used instead 、 and ; Assume that all robots are unaware of the H, C, and g matrices.

[0094]

[0095] is the so-called grasping matrix, which transforms the force / torque exerted by the ith robot at the contact position into an effect denoted by {P};

[0096] The Euler-Lagrangian system (1) has the following properties:

[0097] P1: Inertia matrix is positive definite and bounded.

[0098] P2: Centrifugal Coriolis Matrix It is bounded.

[0099] P3: Matrix-valued functions is obliquely symmetric, that is:

[0100]

[0101] P4: There exists a constant vector and a matrix function , making

[0102]

[0103] in , is called the regressor;

[0104] From the dynamics (1) and expression (6) in P4, we get

[0105]

[0106] in , represents the unknown parameters in the system dynamics equations.

[0107] This invention is oriented to trajectory tracking tasks and designs a decentralized controller for each robot. , without relying on robot-to-robot communication, to carry rigid objects and track the desired trajectory. The desired trajectory is described as

[0108]

[0109] in are the desired position, acceleration, angular velocity and attitude respectively; let , The system and the desired trajectory satisfy:

[0110] Assumption 1: The collaborative handling system - (1) satisfies properties P1-P4;

[0111] Assumption 2: The desired trajectory in is only for the first robot Available; Status and Accessible to all robots.

[0112] 2. Design of Expected Trajectory Estimator and Parameter Estimator

[0113] Further, if Figure 3 As shown in FIG, the structure of the proposed framework, the estimation of the desired trajectory and system parameters, specifically includes:

[0114] The expected trajectory estimator is designed for each helper as follows:

[0115]

[0116]

[0117] in , , and They are Estimates of the desired position, velocity, attitude, and angular velocity of each robot, and is the positive estimated gain; for simplicity, let ;

[0118] The wavy lines on the letters represent the error of the corresponding variable relative to the true quantity, and the corresponding error dynamics equation is obtained:

[0119]

[0120]

[0121] Where A is the Hurvez matrix, s x are the components of the following composite tracking error s.

[0122]

[0123] in It is called the reference speed and is defined as

[0124]

[0125] In the formula represents the posture tracking error;

[0126] Next, the DREM method is used to design parameter estimators for each robot, specifically and As shown below

[0127]

[0128] in is the estimated parameter of the ith robot, is a positive definite matrix, Given in P4, is a positive scalar; is the DREM regressor, which can be derived through the DREM algorithm.

[0129] 3. Controller Design

[0130] Define the estimated reference velocity and the estimated composite error as

[0131]

[0132] Then, the decentralized controllers for all robots are designed as follows:

[0133]

[0134]

[0135]

[0136] in is the formula (2) The inverse matrix of, for i∈V and h,k i is a positive scalar gain; note that in (35)-(36) we use and , to represent and . Define the tracking error and obtain the error dynamics as

[0137]

[0138]

[0139] .

[0140] 4. Proof of Stability

[0141] The Lyapunov function is designed as follows: ,in , is the Lyapunov equation The solution of is a symmetric positive definite matrix. , , , It can be shown that when other gains are given, the gains and and When the value is large enough, trajectory tracking is achieved. In addition, if the regressor When the excitation is continuous, the parameter estimation error converges to zero.

[0142] 1. Specific application fields or related products of the present invention.

[0143] An embodiment of the present invention provides a decentralized collaborative transport system for a dragger and assistor system based on the decentralized collaborative transport method for the dragger and assistor system, the system specifically comprising:

[0144] The desired trajectory estimator module first designs a decentralized estimator for the assistant to estimate the desired trajectory in order to implement control based on it;

[0145] The parameter estimation module uses dynamic regression expansion and hybrid DREM method to estimate unknown parameters, ensuring that the parameters converge to the true value under the condition of continuous excitation;

[0146] The controller module is connected with the desired trajectory estimator module and the parameter estimation module. Based on the desired trajectory estimator and the parameter estimator, a distributed controller of the robot is designed to realize collaborative handling.

[0147] An embodiment of the present invention provides a computer device, which includes a memory and a processor. The memory stores a computer program. When the computer program is executed by the processor, the processor executes the steps of the decentralized collaborative transport method of the dragger-assistant system.

[0148] An embodiment of the present invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the steps of the decentralized collaborative transport method of the dragger-assistant system.

[0149] An embodiment of the present invention provides an information data processing terminal, which is used to implement the decentralized collaborative transport system of the dragger-assistant system.

[0150] 2. Relevant evidence of the technical effects obtained by the embodiments of the present invention.

[0151] In numerical simulations, a group of six robots is used to carry a rigid object to follow a periodic trajectory (8)

[0152]

[0153] The unknown parameters are

[0154]

[0155] set up and In respectively , , , , , , all of which are known quantities. The Euler method is used to solve the dynamics forward, with a simulation step size of 0.001s.

[0156] The filter parameters are λφ1=λφ2=1, λa=1.2 and λb=0.3, and the initial value of the filter is set to zero. The controller gain is = 1500, = 40, the helper's gain is = 100, is the gain of the parameter estimation law, and the gain of DREM is .

[0157] The proposed framework is compared with the work of Culbertson et al. (2021), where each robot knows the required trajectory information and uses the MRAC method. Figure 4 The results show that as time increases, the composite tracking error s converges to 0, which means that and Approaching zero, Approaching In addition, the results of parameter estimation are plotted in Figure 5 middle.

[0158] The physics simulations were performed using the open-source software Gazebo (Koenig and Howard, 2004), which is based on the Robotic Handling System (ROS) (Quigley et al., 2009), which includes packaged libraries and tools for creating robotic applications. The same rigid body model as in Culbertson et al. (2021) was used, as Figure 6 The yellow cylinder marks the object carried by the robot, which is represented by a square. Under the action of forces and torques, the object can achieve linear and rotational motion relative to the cylinder axis. For simplicity, only linear and rotational motion around the cylinder axis are considered. The expected trajectory is

[0159]

[0160] The system settings, including the number of robots, the position of the robots relative to the object, the control troller gain, and the simulation step size are consistent with the numerical simulation. The Euler angle η is used instead of the rotation matrix to represent the rotation posture. The results are as follows Figure 7 As shown. , , The actual Euler angles converge to the expected Euler angles, which shows the effectiveness of the proposed method.

[0161] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware portion can be implemented using dedicated logic; the software portion can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those skilled in the art will appreciate that the above-mentioned devices and methods can be implemented using computer-executable instructions and / or contained in processor control code, for example, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, or can be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.

[0162] The above description is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with this technical field within the technical scope disclosed by the present invention and within the spirit and principles of the present invention should be covered by the scope of protection of the present invention.

Claims

1. A decentralized collaborative transport method for a dragger-assistant system, characterized in that: The method includes: The dynamic modeling of the collaborative handling system consisting of a dragging robot and a helper robot is carried out, including the inertial system, the fixed system fixed to the center of mass of the rigid body, the anchor point and its related position and posture variables. The anchor point is defined as the origin of the auxiliary coordinate system, which is aligned with the direction of the fixed system. The contact torque between the robot and the rigid object is converted into the system dynamic parameters. By setting the vector relationship between the contact position between the robot and the rigid object and the anchor point, the position, velocity, acceleration and posture description are established; The inertia matrix, centrifugal force matrix and gravity vector are used to describe the dynamic parameters, where the inertia matrix is ​​positive and bounded, the centrifugal force matrix is ​​bounded, and meets the matrix characteristic requirements; The regressor form is introduced to express the unknown parameters in the system dynamics equation as the functional relationship between the known regressor and the parameter to be estimated; A decentralized desired trajectory estimator is designed for the helper to estimate the desired trajectory. Dynamic regression extension and hybrid methods are used to estimate the unknown dynamic parameters. Under the condition of continuous excitation, the parameters are guaranteed to converge to the true value. Then, a decentralized controller is proposed to achieve cooperative handling. The design of the desired trajectory estimator includes: Provide an estimator of the expected position, velocity, attitude, and angular velocity for each assisting agent, adjusting the estimation accuracy through a positive gain parameter; By defining the tracking error and calculating the error dynamics based on the reference speed, the dynamic change of the composite tracking error is obtained and used to adjust the control strategy; The convergence condition of the composite tracking error is derived through the error dynamics equation, and the trajectory estimation stability is verified; Using dynamic regression extension and hybrid methods, the unknown parameters of the system dynamics are estimated under continuous excitation conditions, generating an estimated reference velocity and an estimated composite error. Based on the estimated desired trajectory and the estimated unknown dynamic parameters, a decentralized controller is designed for each robot to generate control commands, control the robots to collaboratively carry rigid objects, and achieve desired trajectory tracking. The design of the distributed controller includes: Design controller gain parameters through Lyapunov function and define estimated reference speed and estimated composite error; The regressors and estimated regressors are used in the control instructions respectively, and the estimated parameters are combined to generate the torque control signal applied by each assistor to the rigid body; Under the condition of satisfying the stability of trajectory reference, the control gain is dynamically adjusted to ensure the asymptotic convergence of trajectory tracking error and the zero convergence of dynamic parameter estimation error.

2. A distributed collaborative transport system for a dragger and assistor system based on the distributed collaborative transport method for a dragger and assistor system according to claim 1, characterized in that: The system specifically includes: Expected Trajectory Estimator Module,First, a decentralized expected trajectory estimator is designed for the,assistant to estimate the expected trajectory; The parameter estimation module uses dynamic regression expansion and hybrid methods to estimate the unknown parameters of the dynamics, ensuring that the parameters converge to the true value under the condition of continuous excitation; The controller module is connected with the desired trajectory estimator module and the parameter estimation module. Based on the desired trajectory estimator module and the parameter estimation module, a distributed controller of the robot is designed to realize collaborative handling.

3. A computer device, characterized in that: The computer device includes a memory and a processor, wherein the memory stores a computer program. When the computer program is executed by the processor, the processor executes the steps of the decentralized cooperative transport method of the dragger-assistant system according to claim 1 .

4. A computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of the decentralized cooperative transport method of the dragger-assistant system according to claim 1.

5. An information data processing terminal, characterized in that: The information data processing terminal is used to implement the decentralized collaborative transport system of the dragger-assistant system as claimed in claim 2.

Images