Practical applications of modern controls for booster autopilot design

This paper presents an evaluation of three of the most promising optimal control techniques for use in launch vehicle autopilot design. The methods evaluated are the EASY5 parameter optimizer, H∞, and SANDY/H₂. Both the EASY5 parameter optimizer and SANDY/H₂ are commercially available software packages based on constrained optimization. Recent advances in the application of H, control theory, such as response mode shaping (RMS), have made it a practical method for booster autopilot design. Three designs were successfully completed using a realistic booster model containing 16 bending modes. A summary is given for each method in terms of user friendliness and ability to meet design requirements. The EASY5 parameter optimization routine was the most user friendly and was considered to be an ideal tool for improving an existing design. Recent advances in the application of optimal control make H∞ and SANDY viable tools for booster autopilot design. SANDY was considered the best of the three booster autopilot design methods evaluated because of its flexibility to satisfy most of the design requirements, combined with a constrained parameter optimization