Abstract
Multi-objective particle swarm optimization algorithms encounter significant challenges when tackling many-objective optimization problems. This is mainly because of the imbalance between convergence and diversity that occurs when increasing the selection pressure. In this paper, a novel adaptive MOPSO (ANMPSO) algorithm based on R2 contribution and adaptive method is developed to improve the performance of MOPSO. First, a new global best solutions selection mechanism with R2 contribution is introduced to select leaders with better diversity and convergence. Second, to obtain a uniform distribution of particles, an adaptive method is used to guide the flight of particles. Third, a re-initialization strategy is proposed to prevent particles from trapping into local optima. Empirical studies on a large number (64 in total) of problem instances have demonstrated that ANMPSO performs well in terms of inverted generational distance and hyper-volume metrics. Experimental studies on the practical application have also revealed that ANMPSO could effectively solve problems in the real world.
Highlights
Many-objective optimization problems (MaOPs) are widely applied to many fields in the real world, such as water supply portfolio planning [1], crash safety design of vehicles [2], Mengke Jiang author contributed to this work.To tackle MaOPs, many nature-inspired heuristic algorithms are proposed in recent years (NSGA-II [6], NSGA-III [7], multi-objective particle swarm optimization (MOPSO) [8], and MOEA/D [9])
This paper introduces the R2 contribution and an adaptive strategy into the novel multiobjective particle swarm (ANMPSO) to balance convergence and diversity
Numerical experimental results prove that the proposed ANMPSO can obtain excellent performances in the two sets of MaOP benchmark functions and the practical crash safety design of vehicles
Summary
Many-objective optimization problems (MaOPs) are widely applied to many fields in the real world, such as water supply portfolio planning [1], crash safety design of vehicles [2], Mengke Jiang author contributed to this work.To tackle MaOPs, many nature-inspired heuristic algorithms are proposed in recent years (NSGA-II [6], NSGA-III [7], MOPSO [8], and MOEA/D [9]). This paper introduces the R2 contribution and an adaptive strategy into the novel multiobjective particle swarm (ANMPSO) to balance convergence and diversity. 2. The adaptive strategy that uses the population spacing information to adjust parameters is proposed to improve the distribution of particles with appropriate diversity and convergence.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
