Przeglądaj {{ collection }} wg Autor "Borowska, Bożena"

Teraz wyświetlane 1 - 2 z 2

A Novel Learning Multi-Swarm Particle Swarm Optimization
(Wydawnictwo Politechniki Łódzkiej, 2023) Borowska, Bożena
Particle swarm optimization (PSO) is one of the metaheuristic optimization methods. Because of its many advantages, it is often used to solve real-world engineering problems. However, in case of complex, multidimensional tasks, PSO faces some problems related to premature convergence and stagnation in local optima. To eliminate this inconveniences, in this paper, a new learning multi-swarm particle swarm optimization method (LMPSO) with local search operator has been proposed. The presented approach was tested on a set of nonlinear functions and a CEC2015 test suite. The obtained results were compared with other optimization methods.
Particle swarm optimization: the gradient correction
(Wydawnictwo Politechniki Łódzkiej, 2009) Borowska, Bożena; Nadolski, Stefan
In the following paper, the solution of the optimization problem that relied on the determination of the optimal geometry of two cylindrical coil arrangement evoking magnetic field of specific parameters was presented. The objective of the task was to generate, in a defined active area, the magnetic field of the largest possible gradient and simultaneously keep this gradient relatively constant. The computations were performed using the classical particle swarm optimization as well as the modified method with the correction sensitive to the fitness function gradient introduced to the formula describing the movement of the specific swarm particles. As a result, a considerable enhancement of the optimization process was achieved.