Fig. 7

Influence of the switching strategy and sub-swarms on DOPS performance for the coagulation model. DOPS begins by using a particle swarm search and then dynamically switches (switch region), using an adaptive switching criteria, to the DDS search phase. We compared the performance of DOPS with and without DDS for different sub-swarm searches to quantify the effect of number of sub-swarms and DDS. We used one, two, four, five and eight sub-swarms, with a total of 40 particles divided evenly amongst the swarms. The results presented are the average of \(\mathcal {T}\) = 25 trials with \(\mathcal {N}\) = 4000 function evaluations each. The convergence rates with higher swarm numbers is typically higher but there is no pronounced difference amongst four, five and eight. The multi-swarm with without DDS saturates while DOPS shows a rapid drop due to a switch to the DDS phase