A Tandem Robotic Arm Inverse Kinematic Solution Based on an Improved Particle Swarm Algorithm

Guojun Zhao, Du Jiang, Xin Liu, Xiliang Tong, Ying Sun, Bo Tao, Jianyi Kong, Juntong Yun, Ying Liu, Zifan FangFrontiers in Bioengineering and Biotechnology 2022 · DOI

Abstract

The analysis of robot inverse kinematic (IK) solutions is fundamental to robot control and path planning. Due to the limitations of analytical and geometric methods, intelligent algorithms are increasingly advantageous as they can derive approximate solutions directly from positive kinematic equations. This paper proposes an improved Particle Swarm Optimization (PSO) algorithm for solving robot inverse kinematics. By integrating an adaptive weight adjustment strategy and a position coefficient $k$ in the velocity factor, the algorithm enhances both global and local search capabilities while reducing computation time. Modeling is performed using the Product of Exponentials (POE) method based on spinor theory, which avoids the singularities inherent in traditional Denavit-Hartenberg (DH) modeling. Experimental results across 6-DOF and 7-DOF robotic arms demonstrate that the proposed algorithm achieves near-zero position error and superior convergence speed compared to standard PSO and Quantum PSO (QPSO) variants.

Introduction

Inverse Kinematics (IK) involves calculating the joint angles required to position a robot's end-effector at a specific target pose. While essential for precision tasks, IK is mathematically complex, often involving non-linear equations with multiple or no solutions. Traditional numerical methods frequently struggle with "singularities"—mathematical configurations where the solver fails—and can be computationally prohibitive for real-time applications.

Intelligent algorithms like PSO offer a robust alternative by treating IK as an optimization problem. However, standard PSO often suffers from slow convergence or becomes trapped in local optima. This research addresses these challenges by introducing an improved PSO framework that combines advanced kinematic modeling with adaptive search strategies.

Kinematic Modeling via Spinor Theory

A core innovation of this research is the adoption of the Product of Exponentials (POE) modeling method based on spinor theory. Unlike the widely used DH parameters, which rely on local coordinate systems for each joint, the POE approach describes the motion of a rigid body as a whole relative to a fixed base frame.

Figure 1: Sketch of a general industrial six-degree-of-freedom robotic arm structure showing joint axes and geometric parameters.

As shown in Figure 1, this method simplifies the description of the robot's geometry. By avoiding local coordinate transformations at every joint, the spinor approach inherently bypasses the mathematical singularities that often plague DH-based solvers, providing a more stable foundation for the optimization process.

Improved Particle Swarm Optimization

The proposed algorithm introduces several key enhancements to the standard PSO architecture to improve its precision and efficiency:

  1. Adaptive Weight Adjustment: The setting of weights significantly impacts the algorithm's ability to balance global exploration and local exploitation. This paper implements a strategy where weights adjust dynamically based on the current solution's fitness, preventing the particles from overshooting the target or stalling.
  2. Velocity Factor Optimization: To accelerate computation, a position coefficient $k$ is introduced into the velocity factor, and condition settings based on physical joint limits are applied. This ensures the search remains within the robot's reachable workspace.
  3. Fitness Function Design: The optimization objective is to minimize the error between the actual end-effector position and the target pose.
Figure 2: Schematic diagram of the fitness function, illustrating the distance calculation between the actual position (T) and the target position (T').

As illustrated in Figure 2, the fitness function $f(i)$ calculates the distance between the transformation matrices of the actual and target positions. The goal of the PSO is to drive this distance to zero.

Experimental Validation and Results

The algorithm was tested through three experiments involving a general 6-DOF industrial arm, a PUMA560 arm, and a 7-DOF redundant robotic arm. The proposed method was compared against standard PSO and Quantum PSO (QPSO) across three metrics: accuracy, speed, and convergence.

Precision

The proposed algorithm demonstrated exceptional accuracy. In comparative tests, it achieved a position error of 0 and a maximum orientation error of only $1.29 \times 10^{-8}$ rad. In contrast, the best-performing competing algorithms (PSO and QPSO) showed minimum position errors of $-1.64 \times 10^{-5}$ and orientation errors of $-4.03 \times 10^{-6}$.

Computational Speed

Efficiency is critical for real-time robotics. In the 7-DOF experiment, the proposed algorithm reached a solution in 0.30004 seconds, outperforming the fastest alternative (0.30521s). For the 6-DOF PUMA560, it achieved a time of 0.31851s compared to the next best at 0.33359s.

Convergence Stability

The improved PSO achieved faster and more stable convergence. While standard PSO often required over 350 iterations and exhibited significant fluctuations, the proposed algorithm typically converged within 100 iterations, maintaining its performance even as the complexity of the robotic system increased.

Conclusion

By combining spinor-based POE modeling with an adaptive PSO algorithm, this research provides a high-precision and computationally efficient solution for robot inverse kinematics. The method's ability to maintain accuracy across different robot configurations (6-DOF and 7-DOF) suggests strong potential for multi-arm coordination and complex path planning. Future research will focus on extending this algorithm to real-time trajectory tracking for moving targets.

Summary updated

Abstract

The analysis of robot inverse kinematic solutions is the basis of robot control and path planning, and is of great importance for research. Due to the limitations of the analytical and geometric methods, intelligent algorithms are more advantageous because they can obtain approximate solutions directly from the robot’s positive kinematic equations, saving a large number of computational steps. Particle Swarm Algorithm (PSO), as one of the intelligent algorithms, is widely used due to its simple principle and excellent performance. In this paper, we propose an improved particle swarm algorithm for robot inverse kinematics solving. Since the setting of weights affects the global and local search ability of the algorithm, this paper proposes an adaptive weight adjustment strategy for improving the search ability. Considering the running time of the algorithm, this paper proposes a condition setting based on the limit joints, and introduces the position coefficient k in the velocity factor. Meanwhile, an exponential product form modeling method (POE) based on spinor theory is chosen. Compared with the traditional DH modeling method, the spinor approach describes the motion of a rigid body as a whole and avoids the singularities that arise when described by a local coordinate system. In order to illustrate the advantages of the algorithm in terms of accuracy, time, convergence and adaptability, three experiments were conducted with a general six-degree-of-freedom industrial robotic arm, a PUMA560 robotic arm and a seven-degree-of-freedom robotic arm as the research objects. In all three experiments, the parameters of the robot arm, the range of joint angles, and the initial attitude and position of the end-effector of the robot arm are given, and the attitude and position of the impact point of the end-effector are set to verify whether the joint angles found by the algorithm can reach the specified positions. In Experiments 2 and 3, the algorithm proposed in this paper is compared with the traditional particle swarm algorithm (PSO) and quantum particle swarm algorithm (QPSO) in terms of position and direction solving accuracy, operation time, and algorithm convergence. The results show that compared with the other two algorithms, the algorithm proposed in this paper can ensure higher position accuracy and orientation accuracy of the robotic arm end-effector. the position error of the algorithm proposed in this paper is 0 and the maximum orientation error is 1.29 × 10 –8 . while the minimum position error of the other two algorithms is −1.64 × 10 –5 and the minimum orientation error is −4.03 × 10 –6 . In terms of operation time, the proposed algorithm in this paper has shorter operation time compared with the other two algorithms. In the last two experiments, the computing time of the proposed algorithm is 0.31851 and 0.30004s respectively, while the shortest computing time of the other two algorithms is 0.33359 and 0.30521s respectively. In terms of algorithm convergence, the proposed algorithm can achieve faster and more stable convergence than the other two algorithms. After changing the experimental subjects, the proposed algorithm still maintains its advantages in terms of accuracy, time and convergence, which indicates that the proposed algorithm is more applicable and has certain potential in solving the multi-arm inverse kinematics solution. This paper provides a new way of thinking for solving the multi-arm inverse kinematics solution problem.
BibTeX
@inproceedings{zhao2022tandem,
  author = {Guojun Zhao and Du Jiang and Xin Liu and Xiliang Tong and Ying Sun and Bo Tao and Jianyi Kong and Juntong Yun and Ying Liu and Zifan Fang},
  title = {{A Tandem Robotic Arm Inverse Kinematic Solution Based on an Improved Particle Swarm Algorithm}},
  booktitle = {Frontiers in Bioengineering and Biotechnology},
  year = {2022},
  doi = {10.3389/fbioe.2022.832829},
}
Figure