Robust and Adaptive Algorithms for Backlash and Friction Compensation in Precision Weapon Pointing Systems
Small Business Information
Scientific Systems Company, (Currently Scientific Systems Company, Inc.)
500 West Cummings Park, Suite 3000, Woburn, MA, 01801
Raman K. Mehra
AbstractFuture weapon platforms will require highly precise pointing and tracking capabilities in the presence of unpredictable terrain motions, unmodeled gun barrel flexure, nonlinear and time-varying parasitics such as backlash, friction. saturation and recoil shocks etc. It is thus necessary to develop controllers that achieve rapid and precise tracking capabilities in presence of all these hard nonlinearities and unmodeled dynamics. The proposed effort will be directed towards developing controllers and analyzing the possible improvements in performance that can be achieved for the gun turret system using the following three approaches: (a) Robust Adaptive Nonlinear Control (RANC), (b) Nonlinear Model Predictive Control (NMPC) and (c) Nonlinear Robust Servo Control (NRSC). The key advantages of these approaches are: RANC - using this approach one can directly and non-conservatively incorporate the qualitative knowledge of the structure of the nonlinearities (such as backlash) into the design, even without accurate knowledge of the model parameters; NMPC - in this framework it is possible to handle the known nonlinearities, constraints in states and inputs (e.g. saturation, limit on rate of variations etc.) and time variations of the system; NRSC - this approach provides a methodology' for control design to achieve accurate tracking in the presence or significant modeling uncertainty. Specific Phase I tasks are: (1) Modeling of the system nonlinearities and the overall system dynamics, (2) Robust Adaptive Nonlinear Control. design, (3) Nonlinear Model Predictive Control design, (4) Nonlinear Robust Servo Control design, and (5) Testing and comparison of these three approaches on a simulated model of the gun turret system. Prof. S. Sastry from UC Berkeley and Dr. S. Jain from Polytechnic University Brooklyn will serve as consultants on the project. The commercial applications of this technology are enormous and exist in areas such as the production of optical instruments and devices. electronics industry, high precision machine tools, and other computer controlled machinery such as robotic manipulators.
* information listed above is at the time of submission.