Angela Zhu, Acumen: Simulation of Mechanical System

Most state-of-the-art modeling and simulation tools use a component-based paradigm since a real world system can get very complicated. Tools such as Matlab/Simulink have won great popularity in the control system design community. Simulink is designed as a block-diagram based package and has been extending itself by adding toolboxes to meet demand of current modeling requirements. Due to its add-by-need feature and its commercial-oriented nature, Simulink has very limited extensibility and flexibility. The modeling process with Simulink thus is hard to optimize and could become very clumsy at certain point. Also, this modeling approach does not yield an explicit mathematical formulation of the system which can be used in system stability analysis.

We propose a new modeling method which supports explicit mathematical model of physical systems. A system is considered to be a collection of interacting components. Each component with continuous-time behavior is described by a set of differential equations, and each component with discrete-time behavior is described in functional reactive paradigm. In this way, system components are described with an unified formulation which is natural to its behavior. Moreover, the component equations can be combined together by shared variables to form a system equation. System behavior can be therefore simulated by solving the system equation using numerical methods. System characteristics such as stability can be examined by applying analysis on the system equation.