Trajectory planning method for robot system with extended axes

By employing a time-optimal trajectory planning algorithm based on reachability analysis in an articulated robot system with extended axes, a built-in trade-off parameter mechanism and motion allocation strategy are established. This solves the problems of high complexity of dynamic models and difficulty in trajectory planning in existing technologies, and enables flexible adjustment and efficient calculation of trajectories under different tasks.

CN121492027BActive Publication Date: 2026-11-10SOUTH CHINA UNIV OF TECH
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202511727728.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-11-24
Publication Date
2026-11-10
Estimated Expiration
2045-11-24

AI Technical Summary

Technical Problem

Existing articulated robot systems with extended axes lack the ability to adjust the ratio of cruising motion to time-optimal motion in time-optimal trajectory planning. This results in jitter affecting tracking accuracy in high-precision tasks, while limiting system efficiency in high-efficiency tasks. Furthermore, the construction of dynamic models is complex and computationally expensive.

Method used

A time-optimal trajectory planning algorithm based on reachability analysis is used to build a built-in trade-off parameter mechanism, reconstruct the controllable set of parameter space, and decouple the planned trajectory and extended axis motion of the articulated robot through motion allocation strategy, thereby reducing computational complexity and achieving a dynamic balance between trajectory smoothness and operating efficiency.

Benefits of technology

It enables flexible adjustment of trajectory tracking accuracy and running speed under different task requirements, reduces computational complexity, improves the robustness and adaptability of the robot system, and adapts to the switching between efficient motion and high-precision motion.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121492027B_ABST
    Figure CN121492027B_ABST
Patent Text Reader

Abstract

The application discloses a trajectory planning method of a robot system with an extended axis, and comprises the following steps: S1, determining a constraint trajectory executed by a joint robot in the robot system; S2, constructing a dynamics model of the joint robot, and converting the dynamics model into linear constraints in a parameter space; S3, building an internal trade-off parameter mechanism based on a time-optimal trajectory planning algorithm of reachability analysis; S4, reconstructing a controllable set of the parameter space by using the internal trade-off parameter mechanism; S5, solving the reconstructed controllable set by using a trajectory planning algorithm, so as to obtain a speed and acceleration sequence of the trajectory of the joint robot in the parameter space; and S6, adopting a motion distribution strategy to decouple the constraint trajectory of the joint robot and the motion of the extended axis in the robot system, so as to complete the trajectory planning. The method can meet the requirements of different tasks on trajectory tracking accuracy, running speed and calculation speed, and also reduces the calculation complexity.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to robotics technology, and more specifically to a trajectory planning method for a robot system with an extended axis. Background Technology

[0002] With the widespread application of robots in industrial assembly lines, cargo handling, and other manufacturing fields, higher demands are being placed on the motion performance of robot systems. Therefore, in engineering, robots are often mounted on linear axes to form articulated robot systems with extended axes. This configuration allows the robot system to have a larger workspace, better motion dexterity, and the ability to avoid singularities. In this way, the robot can perform tasks within a larger work area, no longer limited to a single fixed installation position. Currently, there are many optimization objectives for industrial robot motion optimization, among which time-optimal is the primary one. Time-optimal trajectory planning is a key technology that enables the robot system to traverse a predetermined trajectory in the shortest time, under the premise of physical constraints such as robot torque, jerk, and acceleration. Existing time-optimal trajectory planning algorithms for articulated robot systems with extended axes lack the ability to adjust the ratio of cruising motion to time-optimal motion. This leads to significant limitations in practical applications of the time-optimal mode: in high-precision tasks (such as precision assembly), time-optimal trajectories may produce jitter, affecting tracking accuracy; while in high-efficiency tasks (such as logistics sorting), a fixed cruising ratio limits the system's operating efficiency. Therefore, the main research objective of this paper is to propose a trajectory planning algorithm for articulated robot systems with extended axes, which can flexibly adjust the ratio of cruising motion to time-optimal motion in the planned trajectory, so as to meet the needs of industrial robots in different tasks.

[0003] Directly solving time-optimal trajectory planning is a difficult task. Therefore, path parameterization methods are needed to transform motor torque constraints into parameterized acceleration constraints. This involves describing the dynamic model using equations and using the shortest possible time as the objective function. Optimization yields robot motion parameters, ensuring the robot's joints operate under critical saturation torque conditions, thereby maximizing running speed and ultimately minimizing running time. This problem can be broadly categorized into two main methods: one is the numerical integration method, based on the Pontryagin minimum principle. This method constructs a Bang-Bang controlled trajectory on the phase plane, selects the limiting acceleration, and uses numerical integration to solve for the maximum speed the robot can provide under any posture, thus achieving the goal of solving the time-optimal robot problem. However, this method requires traversing the trajectory during planning and calculating the maximum scalar velocity, scalar acceleration, and actual velocity at each point in the phase plane under constraints. Therefore, an accurate dynamic model is needed, which is difficult to construct for an articulated robot system with extended axes, and consumes significant computation time during planning. Therefore, applying motion assignment methods to reduce computational complexity is necessary.

[0004] For trajectory planning algorithms with adjustable running time, it is necessary to weigh the weights of feasible solutions (cruising trajectories) and optimal solutions (time-optimal trajectories) based on the time-optimal trajectory to flexibly adjust the trajectory running time, thereby achieving a dynamic balance between trajectory smoothness and operational efficiency. Typically, this is achieved by considering two mixed objective functions with opposite effects in the planning algorithm: one proportional to the execution time and the other proportional to the integral of the square of the jerk, thus striking a suitable balance between speed and smoothness. However, the trajectory planned by this algorithm still consists of acceleration and deceleration motions, and the robot vibrations excited by acceleration and deceleration cannot better guarantee the accuracy of the end effector. Therefore, a trajectory planning algorithm suitable for articulated robot systems with extended axes is proposed, which is necessary to improve the robustness of the algorithm while ensuring trajectory smoothness and operational efficiency. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a trajectory planning method for a robot system with an extended axis. This trajectory planning method can meet the requirements of different tasks for trajectory tracking accuracy, running speed, and calculation speed, while reducing computational complexity.

[0006] The objective of this invention is achieved through the following technical solution: A trajectory planning method for a robot system with an extended axis, comprising the following steps:

[0007] S1. Determine the constraint trajectory executed by the articulated robot in the robot system;

[0008] S2. Construct a dynamic model of the articulated robot and transform this dynamic model into a linear constraint in the parameter space;

[0009] S3. A time-optimal trajectory planning algorithm based on reachability analysis is built with a built-in trade-off parameter mechanism.

[0010] S4. Reconstruct the controllable set of the parameter space using a built-in trade-off parameter mechanism;

[0011] S5. The reconstructed controllable set is solved using a trajectory planning algorithm to obtain the velocity and acceleration sequence of the articulated robot's trajectory in the parameter space.

[0012] S6. A motion allocation strategy is adopted to decouple the constrained trajectory of the articulated robot from the motion of the extended axis in the robot system in order to complete the trajectory planning.

[0013] Preferably, the constraint trajectory in step S1 is a continuous trajectory generated based on the workspace of the articulated robot that does not contain singularities.

[0014] Preferably, the dynamic model in step S2 is:

[0015]

[0016] Among them, q, and These are the joint angle, velocity, and acceleration vectors, respectively. Represents the inertia matrix. Represent the Coriolis term and the eccentric term.

[0017] This represents the torque vector produced by gravity. This represents the diagonal matrix of Coulomb friction coefficients. This is the joint torque vector.

[0018] Preferably, the specific process of transforming this dynamic model into linear constraints in the parameter space in step S2 is as follows:

[0019] S21. Specify the motion trajectory p of the end effector, which includes the position and attitude information of the end effector in Cartesian space.

[0020] S22. Transform the Cartesian space into a joint space parameter path q using inverse kinematics. d ;

[0021] S23. Introduce path parameter s to transform the geometric information in joint space into a parameterized path q related to s. d(s), s∈[0,L], where s represents the path along p from the starting point of the end effector, and L is the total length of the path;

[0022] S24, let q d (s) is considered as a function with path parameter s, q d (s(t)), and Let the first and second derivatives with respect to s be expressed as follows:

[0023]

[0024] Among them, s, and These represent the position, velocity, and acceleration in the parameter space, respectively.

[0025] S25. Applying the first and second derivatives to joint angle constraints and joint acceleration constraints, we obtain:

[0026]

[0027] in,

[0028] m(s)=M(q d (s))q d′ (s)

[0029] c(s)=M(q d (s))q d "(s)+C(q d (s),q′ d (s))q d ′(s)

[0030] g(s)=τ f (q d (s))sign(q d ′(s))+G(q d (s))

[0031] and τ These represent the upper and lower limits of τ, respectively, and the data format is consistent with that of τ.

[0032] Preferably, step S3 includes the following steps:

[0033] S31. Discretize the parameterized path s∈[0,L] and further divide the constrained trajectory into N grids, with interval endpoints s0, s1, ... s2. N-1 and s N , will s i Considered as the i-th stage, Δ i =si+1 -s i >0 represents the interval. Let be the pseudo-velocity squared value for stage i. Let be the pseudo-acceleration value in the i-th stage. Then the relationship between the state values ​​of adjacent intervals is:

[0034]

[0035] The slope of the velocity curve is:

[0036]

[0037] Admissible control states that satisfy dynamic constraints within a discrete framework. The set is:

[0038]

[0039] Then the set of allowable state values ​​that satisfy the dynamic constraints in the i-th stage is:

[0040]

[0041] The admissible control set for a given state x in stage i that satisfies dynamic constraints is:

[0042]

[0043] S32. Determine the trade-off parameters based on user task requirements. This trade-off parameter The interval is:

[0044]

[0045] Preferably, step S5 includes the following specific steps:

[0046] S51. After reconstructing the controllable set, the trajectory planning algorithm is used in each stage i to select the control with the maximum acceleration. So that the next state to be transitioned to belongs to the controllable set of stage i+1;

[0047] S52. Use reverse iterative calculation to obtain the final velocity to reach the final point;

[0048] S53. Select the maximum permissible state value through forward iteration to generate the time-optimal trajectory.

[0049] Preferably, step S52 includes the following specific steps:

[0050] S521, Let the control set in the (i+1)th stage be... It is known that the goal of the reverse iterative single-step computation is to obtain the control set for the i-th stage.

[0051] S522, Single-step state set of stage i It is a state; a state exists. And a permissible control Make the system from arrive The constraints are satisfied; the controllable set in stage i is represented recursively as follows:

[0052]

[0053] S523, Control Set any state value After several positive permissible driving values Adjustments are made to obtain the final speed at the destination.

[0054] Step S53 includes the following specific steps:

[0055] S531, The set of permissible driving values ​​in the i-th stage is:

[0056]

[0057] S532, the state value in the i-th stage is Furthermore, provided that the torque-related constraints of the j-th joint are satisfied, the key factor is to select the largest possible driving value for the i-th stage. Then the state value at stage i+1 is contained in Under the given conditions, the acceleration sequence can be obtained:

[0058]

[0059] Seek again Then, the optimal state value for the (i+1)th stage can be obtained through the constraint relationships. That is, velocity sequence:

[0060]

[0061] Preferably, step S6 includes the following specific steps:

[0062] S61. Based on the installation position of the articulated robot and the extended axis, let the direction of motion of the extended axis be the angle between the x and y axes of the articulated robot's base coordinate system;

[0063] S62. Let the direction vector be used to decompose the trajectory points of the articulated robot in Cartesian space into components that move along linear guides and project the trajectory points p in the Cartesian space onto the component d.

[0064] S63. Distribute the motion of the articulated robot's workspace to the extended axis, obtaining the linear motion along the extended axis and the motion trajectory of the articulated robot. Project the trajectory point p in the Cartesian space series onto component d:

[0065]

[0066] Thus, the trajectory of the extended axis and articulated robot after motion distribution is obtained:

[0067]

[0068] Among them, P axis P represents the trajectory of the extended axis after allocation. ur To assign motion trajectories to the articulated robot, p i To discretize a specified motion trajectory p into the i-th stage of N networks.

[0069] The present invention has the following advantages over the prior art:

[0070] 1. The trajectory planning method for robot systems with extended axes of the present invention uses a time-optimal trajectory planning algorithm based on reachability analysis to build built-in trade-off parameters, and then uses the built-in trade-off parameter mechanism to reconstruct a controllable set of parameter space, enabling users to flexibly adjust the ratio of time-optimal to cruising motion to meet the needs of different tasks for trajectory tracking accuracy, running speed, and calculation speed. This achieves a trade-off between trajectory requirements for different tasks and facilitates the robot system to switch between efficient and high-precision motion to adapt to different task requirements, providing a new solution for efficient trajectory planning of robot systems with extended axes.

[0071] 3. The trajectory planning method for robot systems with extended axes of the present invention utilizes a motion allocation strategy to decouple the planned trajectory of the articulated robot from the motion of the extended axis. Only the motion vector of the extended axis is needed to complete the trajectory planning, thereby reducing computational complexity. This solves the problem that it is difficult to construct a complete and accurate dynamic model for robot systems with extended axes, which leads to difficulties in optimal trajectory planning for the robot. Attached Figure Description

[0072] Figure 1 This is a flowchart of the trajectory planning method for a robot system with an extended axis according to the present invention.

[0073] Figure 2 This is a structural diagram of an articulated robot that implements the trajectory planning method for a robot system with an extended axis according to the present invention.

[0074] Figure 3 This is a framework diagram of the parameter space reconstruction mechanism built into the present invention.

[0075] Figure 4 This is a schematic diagram of the robot system of the present invention. Detailed Implementation

[0076] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0077] like Figure 2 and Figure 4 As shown, the robot system with an extended axis in this embodiment mainly consists of an articulated robot and an extended axis.

[0078] like Figure 1 As shown, the trajectory planning method for this robot system with an extended axis includes the following steps:

[0079] S1. Determine the constraint trajectory executed by the articulated robot in the robot system; the constraint trajectory in step S1 is a continuous trajectory generated based on the workspace of the articulated robot that does not contain singular points. Singular points refer to configurations where the Jacobian matrix of the articulated robot loses its full rank during motion, and at these locations, the kinematic characteristics of the articulated robot undergo abrupt changes.

[0080] Specifically, the constraint trajectory is generated based on the task performed by the robot system. In this embodiment, the constraint trajectory of the robot system is generated by interpolation from three acquisition points in Cartesian space, let the three acquisition points be A, B and C, and the pose of these three acquisition points includes the position of the X, Y and Z axes and the attitude information of roll angle, pitch angle and yaw angle.

[0081] S2. Construct a dynamic model of the articulated robot and transform this dynamic model into a linear constraint in the parameter space;

[0082] The dynamic model in step S2 is as follows:

[0083]

[0084] Among them, q, and These are the joint angle, velocity, and acceleration vectors, respectively.

[0085] Represents the inertia matrix. Represent the Coriolis term and the eccentric term. This represents the torque vector produced by gravity. This represents the diagonal matrix of Coulomb friction coefficients. This is the joint torque vector.

[0086] The specific process of transforming this dynamic model into linear constraints in the parameter space in step S2 is as follows:

[0087] S21. Specify the motion trajectory p of the end effector, which includes the position and attitude information of the end effector in Cartesian space.

[0088] S22. Transform the Cartesian space into a joint space parameter path q using inverse kinematics. d ;

[0089] S23. Introduce path parameter s to transform the geometric information in joint space into a parameterized path q related to s. d (s), s∈[0,L], where s represents the path along p from the starting point of the end effector, and L is the total length of the path;

[0090] To meet the subsequent computational needs, q d (s) should be at least three-fold continuously differentiable; the path parameter s represents the path along which the end effector starts from the starting point. The path length of the motion, L, is the total path length, and therefore its scalar function with respect to time t is s(t). Without loss of generality, assuming the trajectory starts at t = 0 and terminates at t = T, then we have q d (s(t)), t∈[0,T].

[0091] S24, let q d (s) is considered as a function with path parameter s, q d (s(t)), and Let the first and second derivatives with respect to s be expressed as follows:

[0092]

[0093] Among them, s, and These represent the position, velocity, and acceleration in the parameter space, respectively.

[0094] S25. Transform the joint space dynamics model into a dynamics model with respect to path scalar parameters, that is, apply the first and second derivatives to the joint angle constraints and joint acceleration constraints, to obtain:

[0095]

[0096] in,

[0097] m(s)=M(q d (s))q d ′(s)

[0098] c(s)=M(q d (s))q d "(s)+C(qd (s), q′ d (s))q d ′(s)

[0099] g(s)=τ f (q d (s))sign(q d ′(s))+G(q d (s))

[0100] and τ These represent the upper and lower limits of τ, respectively, and the data format is consistent with that of τ.

[0101] S3. A time-optimal trajectory planning algorithm based on reachability analysis is built with a built-in trade-off parameter mechanism.

[0102] Step S3 includes the following steps:

[0103] S31. Discretize the parameterized path s∈[0,L] and further divide the constrained trajectory into N grids, with interval endpoints s0, s1, ... s2. N-1 and s N , will s i Considered as the i-th stage, Δ i =s i+1 -s i >0 represents the interval. Let be the pseudo-velocity squared value for stage i. Let be the pseudo-acceleration value in the i-th stage. Then the relationship between the state values ​​of adjacent intervals is:

[0104]

[0105] The slope of the velocity curve is:

[0106]

[0107] Admissible control states that satisfy dynamic constraints within a discrete framework. The set is:

[0108]

[0109] Then the set of allowable state values ​​that satisfy the dynamic constraints in the i-th stage is:

[0110]

[0111] The admissible control set for a given state x in stage i that satisfies dynamic constraints is:

[0112]

[0113] S32. Determine the trade-off parameters based on user task requirements. This trade-off parameter The interval is:

[0114]

[0115] S4. Reconstruct the controllable set of the parameter space using a built-in trade-off parameter mechanism; through the built-in trade-off parameters... The upper limit of reconstructable controllable sets is The controllable set is adjusted to a parameterized trajectory with a certain time-optimal cruise ratio. Wherein, given the final velocity... In this case, the algorithm recursively calculates the controllable set. If any controllable set is empty, or the initial velocity is... If a path is not included in the set of order-controllable paths, then it is clearly unrealizable. This step ensures the feasibility of the path and provides a foundation for subsequent trajectory generation.

[0116] S5. The reconstructed controllable set is solved using a trajectory planning algorithm to obtain the velocity and acceleration sequence of the articulated robot's trajectory in the parameter space.

[0117] Step S5 includes the following specific steps:

[0118] S51. After reconstructing the controllable set, the trajectory planning algorithm is used in each stage i to select the control with the maximum acceleration. This ensures that the next state transitioned to belongs to the controllable set of stage i+1; this step guarantees that the generated trajectory, while satisfying the constraints, is as close as possible to the time optimum. To achieve this goal, this paper transforms the time optimum problem into solving N small linear optimization problems. The core of the algorithm is to ensure that the destination is reached at the desired destination speed through directional iterative calculation, i.e., to calculate the controllable set; then, through forward iteration, the maximum allowable state value is selected to generate the time-optimal trajectory.

[0119] like Figure 3 As shown in S52, reverse iteration calculation is used to obtain the final velocity reaching the final point; specifically, the reverse iteration calculation starts from the last grid point N=1000 (i.e., the final point), based on the state value Begin iteratively calculating the controllable set for each stage. Until the calculation at grid point 0 is completed, the optimization objective is to optimize the state values ​​at each stage. Maximum. Controllable set The physical meaning of is the set of state values ​​at stage i, considering a set of desired final states. The controllable set in stage i For state A set of states whose values ​​can be passed through an allowable control sequence. The adjustment to Reaching the final stage Make the system from arrive It is controllable. The reverse iterative calculation consists of multiple single-step processes, and the implementation process is as follows:

[0120] Step S52 includes the following specific steps:

[0121] S521, Let the control set in the (i+1)th stage be... It is known that the goal of the reverse iterative single-step computation is to obtain the control set for the i-th stage.

[0122] S522, Single-step state set of stage i It is a state; a state exists. And a permissible control Make the system from arrive The constraints are satisfied; the controllable set in stage i is represented recursively as follows:

[0123]

[0124] S523, Control Set any state value After several positive permissible driving values Adjustments are made to obtain the final speed at the destination.

[0125] like Figure 3 As shown, S53 generates the time-optimal trajectory by selecting the maximum permissible state value through forward iteration. The forward iteration starts from grid point 0, with the initial state value being... Begin by forward selecting the maximum state value, continuing until the calculation at point N is completed, ensuring that the resulting velocities remain within their respective controllable sets. Consider a set of initial states. Makes existence and allowable control sequence Make the system from arrive Finally in The optimal state value at each stage within the phase plane is the numerical solution of the trajectory that satisfies the physical constraints and has the shortest motion time. Similar to the backward iterative calculation, the forward iterative calculation also consists of multiple single-step processes. The single-step forward iterative implementation process is as follows:

[0126] Step S53 includes the following specific steps:

[0127] S531, The set of permissible driving values ​​in the i-th stage is:

[0128]

[0129] S532, the state value in the i-th stage is Furthermore, provided that the torque-related constraints of the j-th joint are satisfied, the key factor is to select the largest possible driving value for the i-th stage. Then the state value at stage i+1 is contained in Under the given conditions, the acceleration sequence can be obtained:

[0130]

[0131] Seek again Then, the optimal state value for the (i+1)th stage can be obtained through the constraint relationships. That is, velocity sequence:

[0132]

[0133] Specifically, during a single-step calculation, if If the solution cannot be found, the trajectory planning fails; otherwise, the forward iterative process continues until the calculation of the (i+1)th stage is completed.

[0134] S6. A motion allocation strategy is adopted to decouple the constrained trajectory of the articulated robot from the motion of the extended axis in the robot system in order to complete the trajectory planning.

[0135] Step S6 includes the following specific steps:

[0136] S61. Based on the installation position of the articulated robot and the extended axis, let the direction of motion of the extended axis be the angle between the x and y axes of the articulated robot's base coordinate system;

[0137] S62. Let the direction vector be used to decompose the trajectory points of the articulated robot in Cartesian space into components that move along linear guides and project the trajectory points p in the Cartesian space onto the component d.

[0138] S63. Distribute the motion of the articulated robot's workspace to the extended axis, obtaining the linear motion along the extended axis and the motion trajectory of the articulated robot. Project the trajectory point p in the Cartesian space series onto component d:

[0139]

[0140] Thus, the trajectory of the extended axis and articulated robot after motion distribution is obtained:

[0141]

[0142] Among them, P axis P represents the trajectory of the extended axis after allocation.ur To assign motion trajectories to the articulated robot, p i To discretize a given motion trajectory p into the i-th stage of N networks, the specified motion trajectory p is discretized into N grids, with endpoints p0, p1, ..., p1. N-1 and p N , will p i This is considered as stage i.

[0143] As attached Figure 4 As shown in Figure S7, the runnable trajectory of the robot system is generated by sending commands from the control system to the servo driver via the industrial control computer. This controls the movement of the linear drive axis motor, and the rotational motion of the motor is converted into the preceding motion of the articulated robot base platform via a ball screw. The linear drive platform control system uses EtherCAT for real-time communication with the servo driver. The articulated robot communicates in real-time with the robot control cabinet through the control system.

[0144] The proposed algorithm eliminates the need for repeated calculations of the cruise trajectory during the cruise planning phase. Therefore, as the tradeoff parameter decreases (i.e., the cruise ratio increases), the algorithm's computational efficiency gradually improves, demonstrating good real-time performance. Compared to time-optimal trajectory planning algorithms based on reachability analysis, the proposed algorithm reduces computation time by 43.8% when the tradeoff parameter is set to 10% of the maximum speed, without reducing the number of planning points. When the tradeoff parameter is at its maximum value, the computational efficiency of the proposed algorithm is the same as that of time-optimal trajectory planning based on reachability analysis. In addition to improved computational efficiency, the proposed algorithm also has the advantage of flexibly adjusting the trajectory runtime according to task requirements.

[0145] The above-described specific embodiments are preferred embodiments of the present invention and are not intended to limit the present invention. Any other changes or equivalent substitutions made without departing from the technical solution of the present invention are included within the protection scope of the present invention.

Claims

1. A trajectory planning method for a robot system with an extended axis, characterized in that, Includes the following steps: S1. Determine the constraint trajectory executed by the articulated robot in the robot system; S2. Construct a dynamic model of the articulated robot and transform this dynamic model into a linear constraint in the parameter space; the dynamic model in step S2 is as follows: in, , and These are the joint angle, velocity, and acceleration vectors, respectively. Represents the inertia matrix. Represent the Coriolis term and the eccentric term. This represents the torque vector produced by gravity. This represents the diagonal matrix of Coulomb friction coefficients. This is the joint torque vector; The specific process of transforming this dynamic model into linear constraints in the parameter space in step S2 is as follows: S21. Specify the motion trajectory p of the end effector, which includes the position and attitude information of the end effector in Cartesian space. S22. Transform the Cartesian space into a joint space parameter path q using inverse kinematics. d ; S23. Introduce path parameter s to transform the geometric information of the joint space into a parameterized path related to s. Where s represents the path along p from the starting point of the end effector, and L is the total length of the path; S24, Order It can be viewed as a function with a path parameter s. , and Let the first and second derivatives with respect to s be expressed as follows: in , and These represent the position, velocity, and acceleration in the parameter space, respectively. S25. Applying the first and second derivatives to joint angle constraints and joint acceleration constraints, we obtain: in, and They represent The upper and lower limits, and the data format is consistent with... Maintain consistency; S3. Establish a built-in tradeoff parameter mechanism for the time-optimal trajectory planning algorithm based on reachability analysis; Step S3 includes the following steps: S31. Discretize the parameterized path s∈[0,L] and further divide the constrained trajectory into N grids, with interval endpoints s0, s1, ... s2. N-1 and s N , will s i Considered as stage i, For intervals, Let be the pseudo-velocity squared value for stage i. Let be the pseudo-acceleration value in the i-th stage. Then the relationship between the state values ​​of adjacent intervals is: The slope of the velocity curve is: , Admissible control states that satisfy dynamic constraints within a discrete framework. The set is: Then the set of allowable state values ​​that satisfy the dynamic constraints in the i-th stage is: The admissible control set for a given state x in stage i that satisfies dynamic constraints is: ; S32. Determine the trade-off parameters based on user task requirements. This trade-off parameter The interval is: ; S4. Reconstruct the controllable set of the parameter space using a built-in trade-off parameter mechanism; S5. The reconstructed controllable set is solved using a trajectory planning algorithm to obtain the velocity and acceleration sequence of the articulated robot's trajectory in the parameter space. Step S5 includes the following specific steps: S51. After reconstructing the controllable set, the trajectory planning algorithm is used in each stage i to select the control with the maximum acceleration. So that the next state to which the state is transitioned belongs to the controllable set of stage i+1; S52. Use reverse iterative calculation to obtain the final velocity to reach the final point; S53. Select the maximum permissible state value through forward iteration to generate the time-optimal trajectory; S6. A motion allocation strategy is adopted to decouple the constrained trajectory of the articulated robot from the motion of the extended axis in the robot system in order to complete the trajectory planning. Step S6 includes the following specific steps: S61. Based on the installation position of the articulated robot and the extended axis, let the direction of motion of the extended axis be the angle between the x and y axes of the articulated robot's base coordinate system; S62. Let the direction vector be used to decompose the trajectory points of the articulated robot in Cartesian space into components that move along linear guideways. In the component Projection on; S63. Distribute the motion of the articulated robot's workspace to the extended axis, obtaining the linear motion along the extended axis and the motion trajectory of the articulated robot, and then convert the Cartesian space series of trajectory points... In the component Projection on: Thus, the trajectory of the extended axis and articulated robot after motion distribution is obtained: in, This indicates the trajectory of the extended axis after allocation. To assign motion trajectories to the articulated robot, p i To discretize a specified motion trajectory p into the i-th stage of N networks.

2. The trajectory planning method for a robot system with an extended axis according to claim 1, characterized in that, The constrained trajectory in step S1 is a continuous trajectory without singularities generated based on the workspace of the articulated robot.

3. The trajectory planning method for a robot system with an extended axis according to claim 1, characterized in that, Step S52 includes the following specific steps: S521, Order No. Phase control set It is known that the goal of reverse iterative single-step computation is to obtain the nth... Phase control set ; S522, No. Stage single-step state set It is a state; a state exists. and allowable control This makes the system from arrive The constraints are satisfied; the controllable set in stage i is represented recursively as follows: S523, Control Set any state value After several positive permissible driving values Adjustments are made to obtain the final speed at the destination. .

4. The trajectory planning method for a robot system with an extended axis according to claim 1, characterized in that, Step S53 includes the following specific steps: S531, No. The set of permissible driving values ​​at each stage is: S532, in the Phase state value And satisfy the first Given the constraints related to joint torques, the key factor is to select the largest possible first torque. Stage driving value So in the first The state value at each stage is included in Under the given conditions, the acceleration sequence can be obtained: Seek again Then, the first can be obtained through the constraint relationships. Stage optimal state value That is, velocity sequence: 。

Citation Information

Patent Citations

  • Time optimal trajectory planning controller and method combined with iterative learning

    CN110221538A

  • Time optimal trajectory planning method based on flexible joint model

    CN115625708A