Modern commercial geosynchronous telecommunications satellites are multimillion dollar spacecraft consisting of complex hardware assemblies and multiple on-board control systems. Over the past two decades, SC Solutions has developed and implemented custom design, modeling, and optimization tools, including GUI-driven, stand-alone tools for analysis and simulation of various spacecraft operating modes. These tools include detailed models of spacecraft subsystem hardware, such as sensors, actuators, disturbance environment, and 6DOF dynamics of flexible appendages. Our control design tools implement Kalman filter techniques to design spacecraft-specific gains for on-board autonomous control systems. These tools include full nonlinear models of spacecraft hardware and on-board software, and enable linear and nonlinear analysis of these multivariable systems. In addition, our optimization tools have been used to solve both control design and hardware layout problems.
The quadrotor, often called drone, is an ideal platform for a wide range of applications due to its superior mobility and hover capability. We have developed a time-optimal path tracking control algorithm for autonomous operation of a quadrotor. The time-optimal controller enables the quadrotor to track a specified path in the shortest possible time. A trajectory generation algorithm is developed to compute the optimal state trajectories through a series of waypoints in the paths while satisfying constraints on the performance of the system. To accommodate aggressive motions that involve high-speed maneuvers, the controller design is based on a nonlinear dynamic model of a quadrotor.
Our philosophy for aircraft control design is to carry out point design and do gain-scheduling to implement full flight envelope control using a set of explicit model-following multivariable flight control laws. The explicit model takes into account the mission level and flying quality requirements. A control structure is developed that combines explicit model-following methodology with various multivariable control system design approaches such as LQG and H∞. The control methodologies provide for robustness by taking into account various sources of model uncertainty. The design methodology has been applied to various fighter aircraft, short-take-off-and vertical landing (STOVL) aircraft, supermaneuverable aircraft, and hypervelocity vehicles.
Modern aircraft turbofan engines are multivariable systems that require optimal control of steady-state and transient operations. Modern control design techniques such as LQG and H∞ are applied for control of such systems and exploit the truly multivariable nature of these propulsion systems. The engine is modeled as a nonlinear dynamical system using aero-thermodynamic principles. Detailed digital simulations are developed for engine control development. Reduced-order control relevant models are derived for feedback control. The controller is robust to modeling uncertainties, disturbances, and sensor degradation.
Another area of interest is sensor failure detection, isolation, and accommodation. We have developed a unique framework for model-based detection of sensor failures using a bank of Kalman filters that takes model uncertainty and sensor noise into account. The framework has been successfully implemented for multivariable turbofan jet engines. The underlying technology received an R&D 100 award.