Flight Test Comparison of Alternate Strategies for Multi-Loop Control Law Optimization

The optimization of design parameters in a multi-loop set of control laws against relevant specifications can be carried out using one of two approaches: 1) all design parameters can be simultaneously optimized against all the specifications in a "One-Shot" approach or 2) design parameters for the inner-most loop can be optimized against specifications relevant to that loop only and then the inner-most loop fixed and optimization continued with the next loop in a "Nested" approach. Clearly, the "One-Shot" approach yields the most optimal solution as it allows all specifications to interact with all design parameters within a single optimization process. However, the "One-Shot" approach also results in a much more complex optimization problem that is much more time consuming and prone to optimization pitfalls. The "Nested" approach, on the other hand, results in a simpler optimization problem that is less computationally intensive and less prone to optimization issues, but may not result in achieving the best performance from the system. This paper compares and contrasts the two approaches and outlines an optimization strategy for the "Nested" approach that retains the simplicity and computational benefits of the approach, while yielding results that are comparable to the "One-Shot" results in terms of achieving the best performance from the system. Analysis and flight test results, using U.S. Army AFDD's RASCAL helicopter, are presented and discussed. The flight test results provide design guidance for the velocity and position loop performance not available in the current version of ADS-33.