Two Optimized General Methods for Inverse Kinematics of 6R Robots Based on Machine Learning

Abstract

For the 6R robot, there is no analytical solution for some configurations, so it is necessary to analyse inverse kinematics (IK) by the general solution method, which cannot achieve high precision and high speed as the analytical solution. With the expansion of application fields and the complexity of application scenarios, some robots with special configuration have become the research hotspot, and more high-speed and high-precision general algorithms are still being explored and studied. The present paper optimized two general solutions. Elimination is a numerical solution, which has high accuracy, but the solution process is complex and time-consuming. The present paper optimized the elimination method, derived the final matrix expression directly through complex coefficient extraction and simplifying operation, and realized one-step solution. The solving speed was reduced to 15% of the original, and the integrity of the method was supplemented. This paper proposed a new optimization method for the Gaussian damped least-squares method, in which the variable step-size coefficient is introduced and the machine learning method is used for the research. It was proved that, on the basis of guaranteeing the stability of motion, the average number of iterations can be effectively reduced and was only 4-5 times, effectively improving the solving speed.

Inverse Kinematics: An Alternative Solution Approach Applying Metaheuristics

“…The general description of a numerical optimization problem is expressed in (3), where f ( q) is the function to minimize that depends on the vector of design variables q. This function can be subject to inequality and equality constraints g i ( q) and h j ( q), respectively.…”

“…The process of finding the IK is complex [1][2][3], and for this reason, diverse ways to deal with this task have been explored: analytic methods, neural networks, and metaheuristic algorithms. Traditional approximations, as is the case of obtaining the analytical equations via geometric analysis, are limited to working with kinematic chains of 2-, 3-, or, in some cases, 4-DOF at maximum, and their solution requires the use of constraints.…”

Feasibility Study and Porting of the Damped Least Square Algorithm on FPGA

“…Such a combination mainly speeds-up computation and achieves accuracy improvements. Wang et al [34], [35] focused on the number of iterations, adjusting it by means of supervised learning techniques, considering different DoFs and different target platforms for implementation, among which an embedded multi-core system.…”

“…Associated [26] no none 4/6 no N/A [32] no none 4 no SW [33] numerical error 7 no SW [34], [35] machine learning step 6/7 no SW [10] machine learning lambda 8 no SW [36] genetic lambda 6 no N/A [37] analytical lambda 6 no SW THIS WORK no step,lambda 4 yes HW we analyze two different MATLAB implementations that we made available as an open-data [12] of the CERBERO project. We will refer to these implementation as:…”

Automated Inverse Kinematics Configuration Selection for Path Planning of a 6-DOF Robot

“…Aside from classic inverse kinematics solutions, newly published studies have presented numerous different approaches based on machine learning (Wang et al, 2020;Tagliani et al, 2022), the Particle Swarm Optimization (PSO) algorithm (Yiyang et al, 2021), Behavior Tree (Zhang, & Hannaford, 2019), and Artificial Neural Networks (ANN) algorithms (Kshitish et al, 2017;Ahmed et al, 2016;Abderrahim et al, 2023).…”