Improved cuckoo search approach based optimal proportional-derivative parameters for quadcopter flight control

Abstract

ABSTRACT This paper proposes an improved cuckoo search (CS) algorithm for an optimal flight of an unmanned quadcopter using proportional-derivative (PD) controllers. In order to improve the optimisation capability of the standard CS, a new initialisation strategy of nests increases the exploitation search. The choice of an enhanced fitness function consisting of the weighted sum of the integral squared errors for longitude, latitude, altitude and attitude gives better performance. Then, the calculation of the velocity of displacements is inspired by the global nests’ intelligence search or the oriented CS capacity. Results and comparisons to the genetic algorithms, the particle swarm optimisation, the cooperative particle swarm optimisation-cuckoo search and the hybrid neural network PD techniques confirm that PD controllers tuned using the improved CS are efficient for quadrotor full control and trajectory tracking.