摘要
针对神经网络的强自学习性、自适应能力及非线性变换特性,结合陀螺静态漂移误差模型,采用函数型神经网络对捷联陀螺静态漂移误差系数进行了非线性估计,解决了捷联陀螺重复启动时的静态漂移误差系数的在线动态标定问题。
As networks with strong self-learning,self-adapling ability and nonlinear change character,this paper makes nonlinear evalu- ation for strapdown-gyroscope static drift error parameter using function-type nerral networks,solves problem about on-line dynamically calibrating for switch-on to switch-on static drift error parameter of strapdown-switch-on to switch-on static drift error parameter of strap- down-gyroscope.
基金
中国船舶总公司科研基金(编号97J40.5.2)
