Local pursuit as a bio-inspired computational optimal control tool

This work explores the use of a cooperative optimization algorithm, known as “local pursuit”, as a numerical tool for computing optimal control-trajectory pairs. Local pursuit is inspired by the foraging activities of ant colonies and has been the focus of recent work on cooperative control. We present a hybrid approach to numerical optimization that combines local pursuit with existing methods, and solves an optimal control problem in many small pieces, in a manner to be made precise. Our approach can overcome some important limitations of traditional nonlinear programming based techniques, leading to the ability to handle larger problems with the same resources, while avoiding difficulties due to ill-conditioning. The performance of our method is illustrated in a numerical example involving optimal orbit transfer of a simple satellite.