Simulation-Based Assessment of Control Law Robustness (RASCLE)

Modern adaptive and reconfigurable control techniques have the potential to improve greatly aircraft performance following system failures or damage. Certifying adaptive and reconfigurable control laws for use in aircraft is difficult.

A practice commonly used in demonstrating the robustness of aircraft control laws is to insert gain and delay uncertainties in feedback control loops to demonstrate robustness to these uncertainties. This is often labor intensive, requiring an engineer to manually insert each uncertainty of interest and run the simulation to determine its effect.

Barron Associates' RASCLE tool takes a two-pronged approach to improving the current practice for validating control laws. First, it seeks to automate the testing typically done manually by an engineer. Automating the process makes it practical to run a larger number of searches and thus achieve a higher degree of confidence in the system. Second, RASCLE seeks to add “intelligence” to the search process. It does this by closed-form stability proofs for simplified aircraft models but ones that include key nonlinearities. Though these stability proofs do not apply to the full nonlinear simulation, they can lend significant insight to the behavior of the full system and thus guide simulation based analysis.