Automated Generation and Simulation of Component-Based Generalized Bond Graphs

The concept of causality presents a problem for component-based modeling of bond graphs, the uncertainty of the causality of each interaction bond. As a result we use generalized bond graphs instead, which is mathematically represented by its Dirac structure. Through the use of generalized bond graphs we can represent each component as a causal computational model (hybrid-input-output representation) or as an acausal computational model (kernel representation). Our contribution consists of developing a method which exploits the Dirac structure of a generalized bond graph component, and generates both a causal and acausal computational model for that component. A case study, with subsequent simulation results, is used to showcase both methods and help compare and contrast both computational models.