Search Results

As technology for both military and civilian aviation systems mature into a new era, techniques to test and evaluate these systems have become of great interest. To achieve a general understanding as well as save time and cost, the use of computer modeling and simulation for component, subsystem or integrated system testing has become a central part of technology development programs. However, the evolving complexity of the systems being modeled leads to a tremendous increase in the complexity of the developed models. To gain confidence in these models there is a need to evaluate the risk in using those models for decision making. Statistical model validation techniques are used to assess the risk of using a given model in decision making exercises. In this paper, we formulate a transient model validation challenge problem for an air cycle machine (ACM) and present a hardware test bench used to generate experimental data relevant to the model.

In this paper, first an analytical model of a synchronous generator with a pulsed constant power load (CPL) is developed and numerically compared with a detailed simulation model. The analytical model is shown to possess good predictive abilities, thus enabling its use for control purposes. Second, the generator has a proportionalintegral (PI) control inner-loop, whose task is to regulate the generator’s output voltage to a desired reference. A novel outer-loop predictive reference governor (PRG) is designed and tested via simulation. The PRG uses the analytical model to predict the output behavior of the generator over a short time window, and continuously modifies the reference given to the inner-loop in order to maintain stringent steady-state requirements, in spite of demanding power requirements at the CPL. Simulation results illustrate the significant performance advantages of using the PRG versus using the inner-loop PI controller alone.