摘要
将蚁群神经网络应用于生理信息的融合,为提高仿生机器马虚拟速度的预报精度,根据仿生机器马运动器特点,提出了一种蚁群神经网络模型,即蚁群算法和神经网络相结合的方法,通过人体生理信息变化,预测机器马运动的虚拟速度值,实验表明,经过蚁群算法优化后的神经网络比单纯BP神经网络预测精度和收敛速度都有较大提高,而且可以有效避免单纯BP算法容易陷入局部最优的不足,可以在实际中应用。
The article unprecedentedly asserts the application of ant colony optimization neural network in data process of bioinformation. To improve the precision and efficiency for virtual speed prediction of bilmimetic robotic horse,an ant colony optimization neural network model which is the combination of ant colony algorithm and neural network is proposed based on the character of biomimetic robotic horse, and the virtual speed of bilmimetic robotic horse can be predicted via the varying of the human physiology. The experiment demonstrates that compare with the simple BP neural network, the neural network optimized by the ant colony algorithm greatly improved in the predicting accuracy and constringency, furthermore, and it is effective applicable to overcome the local optimization that BP bring about.
出处
《武汉理工大学学报》
CAS
CSCD
北大核心
2009年第13期137-140,共4页
Journal of Wuhan University of Technology
基金
河北省科技厅(07212106D)
关键词
仿生机器马
神经网络
蚁群算法
biomimetic robotic horse
neural network
ant colony algorith
