Mathematical modelling of purely ODE systems by using the Bond Graph technique and taking the inherent causalities

Abstract Simulation and especially computer simulation, is a basic tool since it enables engineers to understand how systems work without actually needing to see them. They can learn how they work in different circumstances and optimize their design with considerably less cost in terms of time and money than if they had to carry out tests on a physical system. In the presented work, an automatic procedure for reducing a system of algebraic-differential equations to a purely differential one, i.e. the minimum number of equations, within a simulation model carried out with a bond graph, and based only on causal assignation, is presented. Depending on the different types of causal paths and algebraic loops coexisting, through a succession of algebraic operations carried out on matrices, the method is capable of obtaining a system of reduced equations. One advantage of this approach is the ease with which the matrix simplification can be programmed by means of a series of operations and derivations, which is especially interesting when it comes to generating symbolic equations for a bond graph model, once they have been reduced and simplified. In each ZCP case will firstly develop the different algorithms and it will then applied to a model, all of this using the Bond Graph technique.