Abstract

The dynamic modeling of multibody systems has been recognized as a key aid in the analysis, design, optimization, and control of mechanical systems. During the past three decades, many software tools, whether commercial or in scientific research, have been introduced that enable the engineers to construct accurate models of multibody systems. These tools introduced powerful methods of the numerical solution, simulation and mathematical transformation that facilitated the process of modeling complex systems. One of the most challenging topics in this context, is the parameter estimation of multibody systems based on laboratory and experimental data and optimal design of such systems. This paper proposes an approach of utilizing optimization methods in the design of multibody systems in which the coordinates-dependent sensitivity of the system is at a definite level with respect to the parameter-dependent sensitivity. In order to attain this condition, the desired dynamic characteristics are introduced instead of experimental data on the estimation process. In the case of gyroscopic systems, these characteristics are either constant or limited in motion, and thus, the sensitivity is highly depended on the system parameters. Consequently, the design of the necessary system parameters can be accomplished with minimal computational effort. The experimental verification of the proposed approach was carried out on the Gyroscopic laboratory system. Eventually, the experimental data of the designed system was tested and compared with the referenced dynamic behavior, which showed acceptable performance in terms of accuracy and efficiency.

摘要

多体系统的动态建模已被认为是机械系统分析、设计、优化和控制的关键辅助工具。在过去的三十年中，引入了许多软件工具，无论是商业软件工具还是科学研究软件工具，使工程师能够构建多体系统的精确模型。这些工具引入了强大的数值求解、模拟和数学变换方法，促进了复杂系统建模过程。在这方面最具挑战性的主题之一是基于实验室和实验数据的多体系统参数估计以及此类系统的优化设计。本文提出了一种在多体系统设计中利用优化方法的方法，其中系统的坐标相关灵敏度相对于参数相关灵敏度处于确定水平。为了达到这个条件，引入了所需的动态特性，而不是估计过程中的实验数据。在陀螺系统的情况下，这些特性要么是恒定的，要么在运动中受到限制，因此，灵敏度在很大程度上取决于系统参数。因此，可以用最少的计算工作完成必要系统参数的设计。所提出方法的实验验证是在陀螺仪实验室系统上进行的。最终，