摘要
将Logistic模型的参数估计问题转化为一个多维无约束函数优化问题，然后利用粒子群优化算法（PSO）的全局搜索能力对此问题求解.仿真实验中所使用的数据包括真实数据和随机采样数据.实验结果表明，在这两种数据条件下PSO算法均能够较准确地估计获得Logistic模型的参数，证实了PSO算法是Logistic模型参数估计的一种可靠有效的算法.同时也分析了参数维数和噪声对PSO算法的收敛性和稳定性的影响.
The parameter estimation for Logistic model was formulated as a multi-dimensional unconstrained function opti- mization problem, and then particle swarm optimization （PSO） algorithm was adopted to solve this problem for its global searching ability. Experimental data includes real-life observation series, randomly sampling data. Experimental results show that PSO algo- rithrn can both obtain quite accurate parameter estimation of Logistic model, and PSO is a reliable and effective method in parameter estimation for Logistic model. Furthermore, the effect of parameter dimension and noise on the stability and convergence of PSO al- gorithm was analyzed as well.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2010年第B02期55-59,共5页
Acta Electronica Sinica
基金
国家自然科学基金(No.60773009)
国家863高技术研究发展计划(No.2007AA01Z290)
湖北省自然基金(No.2007ABA009)
