Design of efficient multiobjective binary PSO algorithms for solving multi‐item capacitated lot‐sizing problem

Abstract

In this paper, a multi-item capacitated lot-sizing problem with setup times and backlogging (MICLSP_SB) is addressed with respect to two conflicting objective functions. This nondeterministic polynomial time (NP)-hard problem consists of finding optimal production plans while minimizing, simultaneously, the total cost and the total inventory level. To effectively solve the considered problem, two new versions of multiobjective binary particle swarm optimization are designed. In the first proposed version "standard multiobjective particle swarm optimization" (S-MOBPSO) algorithm, two main contributions are introduced. First, a new particle encoding and decoding method is developed to treat the MICLSP_SB decision variables. Second, the overall violation minimization method and the constraint handling method are combined to effectively handle the problem constraints. To further enhance the overall performance, an improvement procedure that minimizes the capacity constraints violations is developed and embedded into the S-MOBPSO algorithm. This second version is named improved MOBPSO (I-MOBPSO). An illustrative example is presented to explain the application of the proposed algorithms. Five performance metrics are considered to measure and evaluate the effectiveness of the proposed algorithms. Experimental results demonstrate the

Novel multiobjective particle swarm optimization based on ranking and cyclic distance strategy

“…PSO is an SI algorithm based on mutual learning among particles, which has been extensively applied in single-objective optimization owing to its few parameters, simple structure, and fast convergence speed. [18][19][20] In PSO, each particle is an individual in D-dimensional space, and changes its motion speed by learning from the past optimal position (pbest) and the global optimal position of all particles in the population (gbest). In PSO, N particles in the D-dimensional objective space comprise a swarm, in which the position of the ith particle is a vector of D-dimensions denoted by x x x x = ( , , …, )…”

An integrated container terminal scheduling problem with different‐berth sizes via multiobjective hydrologic cycle optimization

“…Execute QC rearrangement for large vessel, otherwise, i j i Flow (15) where the particle j has better fitness than i, rand () is a random generation function to generate a d-dimension vector, and…”

“…Multiobjective optimization algorithms have revealed good advantages in acquiring solution sets efficiently and suffered extensive development. Such as, Ben Ammar et al 15 proposed new versions of multiobjective binary particle swarm optimization, Niu et al 16 gave multiobjective bacterial colony optimization algorithm, Nourmohammadzadeh and Voss 17 declared multiobjective simulated annealing, and so on. These algorithms mostly adopted convergence first and diversity second principle, which may get stuck at an easy‐to‐find part of the Pareto front, especially in problems with disconnected feasible regions.…”

Energy Dispatching Based on an Improved PSO-ACO Algorithm

“…Ji et al [38,39] proposed a dual-surrogate-assisted and multisurrogate-assisted multitasking PSO, which can obtain multiple optimal solutions of expensive multimodal optimization problems at low computational cost. Ammar et al [40] designed two new multiobjective binary PSO algorithms, which efectively solved multi-item capacitated lotsizing problem. Zheng et al [41] proposed a PSO algorithm based on migration learning, which efectively improves its search efciency in the traveling salesman problem and can obtain better routes.…”