Real-time evolution of an embedded controller for an autonomous helicopter

Abstract

Abstract-In this paper we evolve the parameters of a proportional, integral, and derivative (PID) controller for an unstable, complex and nonlinear system. The individuals of the applied genetic algorithm (GA) are evaluated on the actual system rather than on a simulation of it. This makes implicit a formal model identification for the implementation of a simulator. This also calls for the GA to be approached in an unusual way, where we need to consider new aspects not normally present in the usual situations using an unnaturally consistent simulator for fitness evaluation. Although elitism is used in the GAs, no monotonic increase in fitness is exhibited by the algorithm. Instead, we show that the GA's individuals converge towards more robust solutions.

Supervised Control of a Flying Performing Robot Using Its Intrinsic Sound

“…We applied two genetic algorithms (GA) to tune the existing controllers directly on the robot rather than evaluating individuals in simulation. In doing so, we confirmed that this method provided us with more robust controllers [10]. Furthermore, we propose a novel method for localisation and enhancement of control that does not require additional sensors to be added to the lightweight indoor helicopter.…”

“…In our work, we evaluated the individual solutions of the applied genetic algorithm on the actual system rather than on a simulation of it, thus avoiding the reality gap altogether [10].…”

Managing uncertainty in sound based control for an autonomous helicopter

“…We applied two GAs to tune the heading and altitude controllers of the helicopter. Rather than using a simulation of the system, we used the real helicopter to evaluate the fitness of individuals in the GA. We have shown that the GA tuned heading controller evolved towards more robust solutions due to naturally occurring noise in the system [25]. Based on these results we have tuned the altitude controller in a similar way.…”

Mitigating the effect of background noise in sound based helicopter control

“…Rather than using a simulation of the system, we used the real helicopter to evaluate the fitness of individuals in the GA. We have shown that the GA tuned heading controller evolved towards more robust solutions due to naturally occurring noise in the system [11].…”