Selective Generalized Coordinates Partitioning Method for Multibody Systems With Non-Holonomic Constraints

Author(s):  
Ayman A. Nada ◽  
Abdullateef H. Bashiri

The goal of this research work is to extend the method of generalized coordinates partitioning to include both holonomic and nonholonomic constraints. Furthermore, the paper proposes a method for selective coordinates for integration instead of identifying a set of independent coordinates at each integration step. The effectiveness of the proposed method is presented and compared with full-coordinates integration as well as generalized co-ordinates partitioning method. The proposed method can treat large-scale systems as one of the main advantages of multi-body systems.

Author(s):  
Hong-Chin Lin ◽  
Tsung-Chieh Lin ◽  
KwangHae H. Yae

Abstract This paper proposes a special form of the recursive Newton-Euler formulation that satisfies the skew-symmetric property, which is a necessary condition to ensure global convergence in a class of regressor-based identification and adaptive control (Slotine, 1987a & 1987b; Craig, 1987). For general multibody systems, such a special form has been developed in a reduced Euler-Lagrange formulation, but not in the Newton-Euler formulation, which has been very popular in the computational analysis of large scale systems. The paper successfully constructs a pair of inertia and Coriolis-centrifugal matrices for a “skew-symmetric” recursive Newton-Euler formulation, which can be used in both dynamics simulations and control applications.


Author(s):  
Zdravko Terze ◽  
Dubravko Matijasˇevic´ ◽  
Milan Vrdoljak ◽  
Vladimir Koroman

Differential-geometric characteristics and structure of optimized generalized coordinates partitioned vectors for generally constrained multibody systems are discussed. Generalized coordinates partitioning is well-known procedure that can be applied in the framework of numerical integration of DAE systems. However, although the procedure proves to be a very useful tool, it is known that an optimization algorithm for coordinates partitioning is needed to obtain the best performance. After short presentation of differential-geometric background of optimized coordinates partitioning, the structure of optimally partitioned vectors is discussed on the basis of gradient analysis of separate constraint submanifolds at configuration and velocity level when holonomic and non-holonomic constraints are present in the system. While, in the case of holonomic systems, the vectors of optimally partitioned coordinates have the same structure for generalized positions and velocities, when non-holonomic constraints are present in the system, the optimally partitioned coordinates generally differ at configuration and velocity level and separate partitioned procedure has to be applied. The conclusions of the paper are illustrated within the framework of the presented numerical example.


Author(s):  
Martin J. Vanderploeg ◽  
Jeff D. Trom

Abstract This paper presents a new approach for linearization of large multi-body dynamic systems. The approach uses an analytical differentiation of terms evaluated in a numerical equation formulation. This technique is more efficient than finite difference and eliminates the need to determine finite difference pertubation values. Because the method is based on a relative coordinate formalism, linearizations can be obtained for equilibrium configurations with non-zero Cartesian accelerations. Examples illustrate the accuracy and efficiency of the algorithm, and its ability to compute linearizations for large-scale systems that were previously impossible.


Author(s):  
S Mohammad Mirtaheri ◽  
Hassan Zohoor

Based on Lagrangian mechanics, use of velocity constraints as a special set of quasi-velocities helps derive explicit equations of motion. The equations are applicable to holonomic and nonholonomic constrained multibody systems. It is proved that in proposed quasi-spaces, the Lagrange multipliers are eliminated from equations of motion; however, it is possible to compute these multipliers once the equations of motion have been solved. The novelty of this research is employing block matrix inversion to find the analytical relations between the parameters of quasi-velocities and equations of motion. In other words, this research identifies arbitrary submatrices and their effects on equations of motion. Also, the present study aimed to provide appropriate criteria to select arbitrary parameters to avoid singularity, reduce constraints violations, and improve computational efficiency. In order to illustrate the advantage of this approach, the simulation results of a 3-link snake-like robot with nonholonomic constraints and a four-bar mechanism with holonomic constraints are presented. The effectiveness of the proposed approach is demonstrated by comparing the constraints violation at the position and velocity levels, conservation of the total energy, and computational efficiency with those obtained via the traditional methods.


Sign in / Sign up

Export Citation Format

Share Document