一种自适应扩展粒子群优化算法

An Adaptive Extended Particle Swarm Optimization Algorithm

在粒子群优化算法的基础上,首先把粒子群优化算法的速度更新式中的个体最优位置用粒子群中所有个体最优位置的平均值代替,得到扩展粒子群优化算法;然后,建立了加速系数和粒子群中所有粒子的平均适应度与整体最优位置适应度之差的一种非线性函数关系,得到自适应加速系数扩展粒子群优化算法。由于新的算法利用了所有个体最优粒子的信息,并在进化过程中通过建立的非线性时变加速系数自适应地调整“认知”部分和“社会”部分对粒子的影响,从而提高了算法的收敛速度和精度。4个基准测试函数的对比实验结果说明自适应扩展粒子群优化算法的有效性和优良性能。

On the basis of particle swarm optimization,an extended particle swarm optimization is first presented by replacing personal best particle with the average of personal best particles in swarm.Then,an adaptive acceleration coefficients extended particle swarm optimization is proposed by establishing a nonlinear functional relationship between acceleration coefficients and the difference of the average fitness of all particles and the fitness global best particle.The proposed algorithms apply more particles＇ information,and adjust adaptively ＂cognition＂ component and ＂social＂ component by nonlinear time-varying acceleration coefficients,thus improve convergence performance.The experiment results demonstrate that the proposed algorithms are superior to original particle swarm optimization algorithm.