摘要
研究一种不完备信息系统参数辨识插补法.首先将每个输入信号加上时间戳,即将输入信号进行时间顺序排序,使数据在特定的时刻缺失;并采用样条插补法,自适应补充缺失数据,即当缺失一个输出数据时,应用线性样条插值法填补缺失数据,当连续缺失两个或两个以上的数据时,应用三次样条插值法进行补充;在此基础上应用粒子群算法对插补后的不完备信息系统进行参数精确估计;最后,通过MATLAB仿真实验验证了该方法在不同缺失率下,其参数估计的精确性与有效性.
In this issue,a recognition and interpolation method of parameter in an incomplete information system was discussed.Firstly,made the data lost on a certain time by sorting the input signals in the or-der of time after adding timestamp on each of them.The missing message was supplemented adaptively using spline interpolating,which means the data was supplemented adaptively using linear spline inter-polating when one output signal was lost,using the cubic interpolating when two or more output signals were lost.Based on these,the parameter was precisely estimated by particle swarm optimization in the interpolation system.Finally,the precision and effectiveness of the parameter estimation in different missing rates was verified by the method of MATLAB simulation experiment.
出处
《安徽工程大学学报》
CAS
2015年第5期69-77,共9页
Journal of Anhui Polytechnic University
基金
国家自然科学基金资助项目(61203033
61172131
61271377)
关键词
不完备信息
样条插值
粒子群算法
参数估计
incomplete information
spline interpolating
particle swarm optimization
parameter esti-mation
