融合迭代和问题维度的速度约束粒子群算法

Particle Swarm Optimization with Velocity Limit Combining Iteration and Problem Dimension

粒子群算法广泛应用于工程、科学与管理等领域实际问题中的复杂优化问题求解,设计新的策略以应对算法的性能和效率瓶颈是该领域的研究热点。针对传统粒子群算法速度约束策略比较单一,容易导致算法收敛速度慢,性能低等问题,提出一种融合算法迭代和问题维度的速度约束策略。通过分析算法种群进化状态评估值与迭代次数及问题维度的关系,设计计算进化状态评估值的公式,使其受算法迭代次数和问题维度影响,最后根据进化状态评估值计算算法的速度约束范围,得到一种融合迭代和问题维度的速度约束粒子群算法。新的速度约束策略使粒子群算法的种群状态受到迭代次数和问题维度的影响,具有自适应性,并对不同维度问题求解具有扩展性,提高了粒子群算法的收敛速度和求解精度,仿真实验证明了算法的有效性。

PSO is widely used to solve complex optimization problems in practical problems in the fields of engineering,science and management.Designing new strategies to deal with the performance and efficiency bottlenecks of the algorithm is a research hotspot in this field.In order to solve the problem that the original velocity limit strategy of PSO is relatively simple,which may easily lead to slow convergence speed and low performance of the algorithm,this paper proposes a new velocity limit strategy combining iteration and problem dimension.By analyzing the relationship of the algorithm evolutionary state evaluation to iterations and the dimension of problem for particle swarm optimization,a formula was designed to calculate the ESE influenced by the iterations and problem dimension,and calculated the velocity limit on the basis of the ESE,so a particle swarm optimization with velocity limit combining iteration and problem dimension was obtained.Finally,the algorithm was affected by iteration and problem dimensions,adaptive and scalable for solving problems in different dimensions.The results show that the strategy improves the convergence speed and accuracy.Experimental results prove the effectiveness of the algorithm.