摘要
为了提高RBF神经网络的预测效果,针对RBF神经网络的不足提出一种自适应混沌粒子群优化RBF神经网络算法(SACPSO-RBF)。该算法主要采用自适应和混沌操作,采用粒子群算法对RBF神经网络进行优化。利用个体的变异、个体之间的交叉操作和个体的混沌扰动,有效克服了粒子群算法的局部收敛性。使用该优化算法对测试函数序列进行预测,并对算法的有效性进行检验。实验结果表明:改进的算法具有良好的预测能力。
In order to improve the prediction results of RBF neural network,this paper proposed an adaptive chaos particle swarm optimization algorithm of RBF neural network（ SACPSO-RBF） according to the shortcomings of the RBF neural network. This algorithm mainly used the adaptive chaotic particle swarm optimization algorithm to optimize the RBF neural network. The use of variation between individuals,the cross over operation and individual chaos disturbance can effectively avoid the local convergence of the particle swarm optimization algorithm. Using the optimization algorithm,we predicted the testing function sequence and tested the validity of the algorithm. Finally,the experimental results show that the algorithm has well prediction ability.
出处
《重庆理工大学学报（自然科学）》
CAS
2015年第11期126-130,共5页
Journal of Chongqing University of Technology：Natural Science
基金
山东省优秀中青年科学家科研奖励基金资助项目(BS2014DX015)
