A dual population based firefly algorithm and its application on wireless sensor network coverage optimisation

摘要

Firefly algorithm (FA) has shown good performance on many engineering optimisation problems. Recent study has pointed out that FA suffers from slow convergence. To enhance the performance of FA, this paper presents a dual population based FA (called DPFA). In DPFA, the entire population consists of two sub-populations. A memetic FA (MFA) and the standard differential evolution are used to generate new solutions in different sub-populations. To verify the performance of DPFA, we test it on nine benchmark functions. Simulation results show that DPFA outperforms MFA and other improved FA algorithms. Finally, we use the proposed DPFA to solve wireless sensor network coverage optimisation problems. Results show that DPFA can also achieve promising solutions.