HIDMS-PSO with Bio-inspired Fission-Fusion Behaviour and a Quorum Decision Mechanism

Abstract

In this study, we propose a new variant of the HIDMS-PSO algorithm with a bio-inspired fission-fusion behaviour and a quorum decision mechanism (FFQ-HIDMS-PSO). In the new algorithm, units are conceptualised as self-organising fission-fusion societies that determine and adopt a suitable behaviour using unit-based quorum decisions. The incorporation of the two bio-inspired mechanisms provide "diversity aware" selforganising units that react to stagnation of particles by adopting a suitable fission-fusion behaviour, leading to a more efficient algorithm capable of maintaining significantly better population diversity throughout the search. The performance of the proposed algorithm was verified with three distinct experiments conducted using CEC'17 and CEC'05 test suites at 30 and 50 dimensions, comparing against 12 state-of-the-art metaheuristics and 12 state-of-the-art PSO variants. The proposed algorithm showed superior performance in these experiments by outperforming all 24 algorithms in all three experiments at 30 and 50 dimensions. The empirical evidence suggests that the proposed method also maintains significantly superior population diversity in comparison to the original HIDMS-PSO.

HIDMS-PSO Algorithm with an Adaptive Topological Structure

“…For the first experiment, the following algorithms were employed: two inertia weight PSO algorithms with different parametric settings (ω = 0.9 → 0.4, c 1 , c 2 = 2 and ω = 0.4, c 1 , c 2 = 2) and HIDMS-PSO (ω = 0.99 → 0.29, c 1 = 2.5 → 0.5, c 2 = 0.5 → 2.5) [7], metaheuristics ( the bat algorithm (BA) [12], grey wolf optimiser (GWO) [13], butterfly optimisation algorithm (BOA) [14], whale optimisation algorithm (WOA) [15], moth flame optimisation (MFO) [16], artificial bee colony (ABC) [17], invasive weed optimisation (IWO) [18], flower pollination algorithm (FPA) [19] and cuckoo search algorithm (CS) (p = 0.25) [20]). For the second and third experiments, we have employed the following PSO variants: genetic algorithm assisted HIDMS-PSO (GA-HIDMS-PSO) (ω = 0.99 → 0.29, c 1 = 2.5 → 0.5, c 2 = 0.5 → 2.5) [21], HIDMS-PSO with bio-inspired fission-fusion behaviour and a quorum decision mechanism (FFQ-HIDMS-PSO) (ω = 0.99 → 0.29, c 1 = 2.5 → 0.5, c 2 = 0.5 → 2.5) [22], standard HIDMS-PSO (ω = 0.99 → 0.29, c 1 = 2.5 → 0.5, c 2 = 0.5 → 2.5) [7], HPSO-TVAC [23], FDR-PSO [24], heterogeneous comprehensive learning and dynamic multiswarm PSO (HCLDMS-PSO) (ω = 0.99 → 0.29, c 1 = 2.5 → 0.5, c 2 = 0.5 → 2.5, pm = 0.1, V max = 0.5 * Range)) [25], heterogeneous comprehensive learning PSO (HCLPSO) [26], self-organizing hierarchical PSO with jumping time-varying acceleration coefficients (MNHPSO-JTAC) [27], χPSO (ring with neighborhood radius n r = 2, φ = 4.1, χ = 0 : 72984, c 1 , c 2 = 2.05) [28], bare bones PSO (BBPSO) [29], dynamic multi-swarm PSO (DMS-PSO) (ω = 0.729, c 1 , c 2 = 1.49445, V max = 0.5 * Range) [30], fully informed particle swarm optimizer (FIPS) [31], unified PSO (UPSO) [32] and comprehensive learning PSO (CLPSO) (ω = 0.9 → 0.2, c 1 , c 2 = 1.49445, V max = 0.2 * Range) [33] and SRPSO (ω = 1.05 → 0.5, c 1 , c 2 = 1.49445, V max = 0.06708 * Range) [34]. In the conducted experiments, the population of all baseline metaheuristics were set as 100, and the population of PSO variants (including the proposed algorithm) was set to 40 [7].…”