Intelligent Control For Autonomous Remote Spacecraft

We will apply the latest knowledge in learning control, adaptive control, and optimal control to develop a modular, state-of-the-art, adaptive, nonlinear, guidance, navigation, and control package for remote spacecraft. The system will incorporate planning and decision making modules to give the remote spacecraft on-line goal directed self-reliant behavior with a high degree of autonomy. At the highest level, an expert system based Goal Executive will monitor the spacecraft's environment and switch and/or adapt the performance indices of an adaptive critic based guidance system, and an evolutionary algorithm based navigation system. This will, in effect, change the near-term behavior of the remote spacecraft by changing what it considers optimal. The navigation system will specify optimal way-points for achieving the specified mission goals. The guidance module will generate an optimal trajectory between the way-points and the control module will ensure robust and accurate trajectory tracking. The controller will be based on an improved version of an nth order multi-variate Neural Adaptive Controller (NACTM) developed at AAC. ~000118 ~