Application of a Near-Optimal Closed Loop Control Law to a Pursuit- Evasion Game Between Two Spacecraft
Abstract
A near-optimal closed loop control law for zero-sum perfect information differential games is tested on a free final time, minimax range, pursuit-evasion game between two constant thrust spacecraft. The control law is based on a periodic first order update to the costate vector. This costate correction is generated from the state error from a reference TPBVP solution. Two nearly equivalent methods to obtain the relationship between the costate correction and the state error are tested. Three coplanar trajectories and two non-coplanar trajectories are tested using various vehicle characteristics. The control law is shown to provide the pursuer an effective means to take advantage of non-optimal play by his opponent; thereby reducing the final range below the original TPBVP solution. Plots comparing the TPBVP solutions and the trajectories resulting from application of the control law against several non- optimal evaders are presented.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1975
- Accession Number
- ADA019324
Entities
People
- Gary D. Bohn
Organizations
- Air Force Institute of Technology