Time-varying PSO – convergence analysis, convergence-related parameterization and new parameter adjustment schemes

Analysis of Particle Swarm Optimization Algorithms with Multiswarms and Time-Varying Parameters

“…where analyze the existing PSO algorithms with time-varying parameters [13,15] by applying the method proposed in [10]. We consider the following four approaches as typical cases of time-varying parameter settings.…”

“…• In [15], a convergence-related parameter adjustment scheme (CRPAS) has been proposed based on the convergence condition [9] of the PSO algorithm. In the reference, 9 time-varying parameter schemes are investigated in the numerical experiments, and two schemes among them outperforms the others including the LDIWM and TVAC.…”

“…In [15], numerical experiments including 100 trials were performed for each test function under the condition with 50 particles for k max = 1000 iterations in a 5-dimensional search space. Not only CRPAS-S7 and CRPAS-S8 but also LDIWM and TVAC were performed as CRPAS-S1 and CRPAS-S4, respectively, in [15]. Here we employ the experimental results obtained in [15] and show them in Table 2.…”

“…Not only CRPAS-S7 and CRPAS-S8 but also LDIWM and TVAC were performed as CRPAS-S1 and CRPAS-S4, respectively, in [15]. Here we employ the experimental results obtained in [15] and show them in Table 2. We see from the table that CRPAS-S7 and CRPAS-S8 give good performance.…”

“…Another direction is to vary the parameters in the PSO algorithm during the computation process. Although some adjustment schemes of the parameters have been proposed in [13,14,15], such time-varying parameters are not based on quantitative analysis such as [10].…”