摘要
采用遗传学习算法和误差反向传播(BP)算法相结合的混合算法来训练前馈人工神经网络,从而提高神经网络的收敛质量和收敛速度,并将此算法运用到电子鼻对可乐的检测上.与经典BP网络及附加动量项BP网络的训练与预测进行了比较,结果显示:遗传优化BP算法具有预测精度高、收敛速度快及运行时间短的优点,是一种快速、可靠的方法.
The combination of genetic algorithm and back propagation algorithm for training the neural network is described. It can improve the search efficiency and realize global optimization, and this GA-BP algorithm is employed to detect the cola by electronic nose. Compared with the standard back propagation algorithm and its improved method, the result shows the GA-BP algorithm has good prediction precision, high convergent speed and less running time, and it is a fast and credible method.
出处
《传感技术学报》
CAS
CSCD
北大核心
2007年第6期1211-1214,共4页
Chinese Journal of Sensors and Actuators
基金
国家自然科学基金项目资助(3057746)
教育部新世纪优秀人才支持计划项目资助(NET-04-0544)
关键词
可乐
电子鼻
BP神经网络
遗传算法
cola
electronic nose
BP neural network
genetic algorithm
