摘要
用滤波理论讨了几类随机扩散过程参数估计的优化问题，得到了在一定条件下偏差系数参数θ的相合极大似然估计与均方误差及在无偏序贯方案下参数θ的最佳极大似然估计，从而满意地解决了上述随机扩散过程参数估计的最优量化问题。且由于方案，是无偏的，量的分布完全是正态的，因而保证了构成θ的罡信区间的可能性．其方法对于优化随机扩散过程的参数估计具在较大的理论与实际意义．
Making use of filter theory, this article reviews several majorization problems of parameter estimation of random scattering process. Thus under certain condition we can get congruent maximum likelihood estimator of parameter θ of bias tactor and error of mean square, We can also get the best maximum likelihood estirnator of parameter e under unbiased sequential plan So the best measuring problem of parameter estimation of radom seattering process can be settled with satisfactory. Because plan is unbiased and the distribution of is entirely normal, the potential of confideuce interral making θ is ensured. The methods of this article have great significance on theory and practice to majorization of parameter estimation of random scattering process.
关键词
扩散过程
参数估计
最佳化
Stochastic-diffusion processes
Parameter estimations
Optimization
Wiener processes
Sequential estimation
Confidence intervals.
