No abstract

Hydraulic and pneumatic networks are highly nonlinear and difficult to analyze. This study presents a software application designed to help students, visualize and understand fluid systems' dynamic behaviors. The application uses a combined bond graph and singular perturbation approach for system equation formulation. A standard iterative and adaptive integrator provides online numerical solutions to the system equations. Coupled to the integrator's output are a graphical animation subsystem and an instrumentation subsystem. The animation subsystem is responsible for rendering movable components on screen, at every simulation time-step, creating the illusion of continuous movement. The instrumentation subsystem collects and displays numerical data in numerical and graphical forms. An interesting contribution of this fluid system analyzer is its ''user-in-the-loop'' feature. This feature allows students to become active participants by enabling them to interact with network components while a simulation run is in progress.

The optimization of classification systems is often confronted by the solution over-fit problem. Solution over-fit occurs when the optimized classifier memorizes the training data sets instead of producing a general model. This paper compares two validation strategies used to control the over-fit phenomenon in classifier optimization problems. Both strategies are implemented within the multi-objective NSGA-II and MOMA algorithms to optimize a Projection Distance classifier and a Multiple Layer Perceptron neural network classifier, in both single and ensemble of classifier configurations. Results indicated that the use of a validation stage during the optimization process is superior to validation performed after the optimization process.