Heap-based optimizer embedded with search strategies applied to high-order analog filter designs: A comparative study with up-to-date metaheuristics

Abstract

Heap-Based Optimizer (HBO) is one of the recently proposed metaheuristic algorithms inspired by the corporate rank hierarchy. In this study, opposition and Cauchy-based search mechanisms are integrated into HBO which is applied for the design of high-order Butterworth active filters, for the first time, comparing with six outstanding metaheuristics introduced for the last year (2020). In addition, the proposed algorithm has been adapted to solve the benchmark presented in the IEEE-CEC 2020 competition to demonstrate its effectiveness against numerical functions. The analog filter design is a challenging problem due to its discrete topology and complex search space. This study provides a comprehensive review for the design of these filters with the up-to-date algorithms: firstly, the performance of each metaheuristic is analyzed by its error value performance, the required number of iterations to achieve this error value; secondly, a statistical test is conducted to validate its performances; as the third, convergence abilities of algorithms are compared with respect to the total design process error values versus an iteration number graph. The passive component values of the Butterworth active filter are selected within the E24 and E96 industrial series so as to make the realization of designs easier to the real world. To demonstrate the suitability of the components obtained by the algorithms to the real world, the amplitude response of the associated design is given. The simulation results demonstrate the effectiveness of the proposed algorithm over the contestant algorithms, and its capability of solving these case design problems. This extensive work can also provide guidelines to the researchers for future studies.