A Novel Hybrid Quantum Particle Swarm Optimization With Marine Predators for Engineering Design Problems

computational efficiency of the quantum particle swarm optimization (QPSO) is significantly higher than that of the traditional particle swarm optimization when solving the parameters of the optimization problem model. However, the exploration and exploitation abilities of the QPSO are weak, based on the historical best position and the global best position. Enlightened by the multi-stage search strategies of marine predators algorithm, we propose a novel hybrid quantum particle swarm optimization algorithm with marine predators (HMPQPSO) in this paper. The evolutionary process of the algorithm is divided two stages: firstly, the Brownian motion of the predator is introduced to the exploration. The randomness and uniformity of which can expand the solution space of particles; secondly, the Levy motion strategy and dynamic parameter strategy are used to update the position, which can accelerate the convergence of the algorithm and enhance the diversity of the algorithm. Meanwhile, both fish aggregation devices(FADs) and opposition-based learning strategy incorporated are used to increase the diversity of the population and prevent the phenomenon of premature particle aggregation. The algorithm is applied to distinct types of CEC2017 benchmark test functions and four multidimensional nonlinear structure design optimization problems, as compared to other recent algorithms. The results demonstrate that the convergence speed and accuracy of HMPQPSO are notably superior to that of other algorithms.