A Grouping Particle Swarm Optimizer with Personal-Best-Position Guidance for Large Scale Optimization

Abstract

Particle Swarm Optimization (PSO) is a popular algorithm which is widely investigated and well implemented in many areas. However, the canonical PSO does not perform well in population diversity maintenance so that usually leads to a premature convergence or local optima. To address this issue, we propose a variant of PSO named Grouping PSO with Personal- Best-Position (Pbest) Guidance (GPSO-PG) which maintains the population diversity by preserving the diversity of exemplars. On one hand, we adopt uniform random allocation strategy to assign particles into different groups and in each group the losers will learn from the winner. On the other hand, we employ personal historical best position of each particle in social learning rather than the current global best particle. In this way, the exemplars diversity increases and the effect from the global best particle is eliminated. We test the proposed algorithm to the benchmarks in CEC 2008 and CEC 2010, which concern the large scale optimization problems (LSOPs). By comparing several current peer algorithms, GPSO-PG exhibits a competitive performance to maintain population diversity and obtains a satisfactory performance to the problems.

Multi‐swarm competitive swarm optimizer for large‐scale optimization by entropy‐assisted diversity measurement and management

“…The experiments results are given in Table 2. The performances of the peer algorithms are directly cited from References 23 and 34. The p ‐value is obtained by Wilcoxon signed ranks test between the proposed algorithm and corresponding peer algorithm.…”

Adaptive fuzzy control for a marine vessel with time‐varying constraints

“…, from which we know that time-varying output constraints can be guaranteed as t → ∞. In other words, (14) not only guarantees the satisfactory transient response performance, but also prevents the violation of timevarying output constraints.…”

“…In [20,35,36], control design mainly focuses on the transient response performance, but the disadvantage is that as t → ∞, the considered constraint functions corresponding to (14) converge to a constant, this point is very conservative. To overcome this conservatism, we introduce a series of novel continuous constraint function given by (14) and the constraint function (14) will converge to time-varying function ϱ i,j cos(ω i,j t) + π ∞i,j as t → ∞. From engineering standpoint, time-varying constraints are more general than constant constraints.…”

“…For the proposed control, control parameters are chosen as follows: For parameters of time-varying constraint function (14), upper bound parameters are given as follows, initial values π 01,1 = π 02,1 = π 03,1 = 0.01, steady-state values π ∞1,1 = π ∞2,1 = π ∞3,1 = 0.002, attenuation rate ξ 1,1 = ξ 2,1 = ξ 3,1 = 2, amplitude of vibration ϱ 1,1 = ϱ 2,1 = ϱ 3,1 = 0.001, and vibration frequency ω 1,1 = ω 2,1 = ω 3,1 = 8, and lower bound parameters are given as follows, initial values π 01,2 = π 02,2 = π 03,2 = −0.01, steady-state values π ∞1,2 = π ∞2,2 = π ∞3,2 = 0.002, attenuation rate ξ 1,2 = ξ 2,2 = ξ 3,2 = 2, amplitude of vibration ϱ 1,2 = ϱ 2,2 = ϱ 3,2 = −0.001, and vibration frequency ω 1,2 = ω 2,2 = ω 3,2 = 8. Velocity constraint bound b is set as b = 0.0287.…”

“…And aggressive nature changes in the marine environment may result in marine vessel collision accidents, therefore a more accurate controller is of greater importance to the safety of the marine vessel. In recent years, many researchers have tried to employ neural network control [7][8][9][10], optimal control [11][12][13][14], robust control [15,16] etc., to achieve the accurate control of some complicated systems in an intelligent and automatical form. If there exists an adaptive control constraining the motion of a marine vessel in a prescribed region, marine vessel collision accidents may decrease.…”