摘要
提出了一种自适应粒子群神经网络自动识别孵化早期种蛋成活性的方法.通过主成分分析提取孵化种蛋颜色特征,减少了神经网络输入节点数.提出的自适应粒子群优化算法,用于优化多层前馈神经网络的拓扑结构,提高了神经网络的学习质量和速度.实验表明该方法识别种蛋成活性切实可行,识别准确性高,算法具有鲁棒性.
A self-adapted PSO neural network for automatic identifying fertility of hatching eggs is given. The primary components of feature parameters are extracted and selected with primary component analysis (PCA). The structure of multi-layer feedback forward neural network is optimized by improved PSO. Learning quality and training speed of the neural network are improved. The result shows that the neural network model for fertility of hatching eggs detection has high accuracy and efficiency and the algorithm is robust.
出处
《内蒙古大学学报(自然科学版)》
CAS
CSCD
北大核心
2006年第4期464-467,共4页
Journal of Inner Mongolia University:Natural Science Edition
基金
内蒙古自然科学基金(200408020809)