In a canonical particle swarm optimization (PSO) algorithm, the fitness is a widely accepted criterion when selecting exemplars for a particle, which exhibits promising performance in simple unimodal functions. To improve a PSO's performance on complicated multimodal functions, various selection strategies based on the fitness value are introduced in PSO community. However, the inherent defects of the fitness-based selections still remain. In this paper, a novelty of a particle is treated as an additional criterion when choosing exemplars for a particle. In each generation, a few of elites and mavericks who have better fitness and novelty values are selected, and saved in two archives, respectively. Hence, in each generation, a particle randomly selects its own learning exemplars from the two archives, respectively. To strengthen a particle's adaptive capability, a multipleinput multiple-output fuzzy logic controller is used to adjust two parameters of the particle, i.e., an acceleration coefficient and a selection proportion of elites. The experimental results and comparisons between our new proposed PSO, named as MFCPSO in this paper, and other 6 PSO variants on CEC2017 test suite with 4 different dimension cases suggest that MFCPSO exhibits very promising characteristics on different types of functions, especially on large scale complicated functions. Furthermore, the effectiveness and efficiency of the fuzzy controlled parameters are discussed based on extensive experiments.

Vehicle routing problem with time windows (VRPTW), which is a typical NP-hard combinatorial optimization problem, plays an important role in modern logistics and transportation systems. Although the particle swarm optimization (PSO) algorithm exhibits very promising performance on continuous problems, how to adapt PSO to efficiently deal with VRPTW is still challenging work. In this paper, we propose a neighborhood comprehensive learning particle swarm optimization (N-CLPSO) to solve VRPTW. To improve the exploitation capability of N-CLPSO, we introduce a new remove-reinsert neighborhood search mechanism. We calculate an information matrix (IM) recording the probability of adjacency between two clients based on the information about the clients themselves and the local information about the elite individuals to guide the removal operation of the neighborhood search. At the same time, we combine the cost matrix (CM) that records the cost of customer removal with IM to create a guided reinsertion operator based on local information to guide the routing. Moreover, to enhance the exploration of N-CLPSO, a semi-random disturbance strategy is proposed. To prevent degradation of the population, the longest common sequences of elites are saved when performing the disturbance. To illustrate the effectiveness of N-CLPSO, this paper conducts extensive experiments

No abstract