Active Guidance for a Finless Rocket Using Neuroevolution

Abstract

Finless rockets are more efficient than finned designs, but are too unstable to fly unassisted. These rockets require an active guidance system to control their orientation during flight and maintain stability. Because rocket dynamics are highly non-linear, developing such a guidance system can be prohibitively costly, especially for relatively small-scale rockets such as sounding rockets. In this paper, we propose a method for evolving a neural network guidance system using the Enforced SubPopulations (ESP) algorithm. Based on a detailed simulation model, a controller is evolved for a finless version of the Interorbital Systems RSX-2 sounding rocket. The resulting performance is compared to that of an unguided standard full-finned version. Our results show that the evolved active guidance controller can greatly increase the final altitude of the rocket, and that ESP can be an effective method for solving real-world, non-linear control tasks.