估计

By means of proposing a priori Markovian transition probability used in the model predictive probability in input synthesize, the IMM estimation algorithm not only utilizes the sufficiency system's posteriori information, but also utilizes rationally the system's priori information, which enhances estimation performance highly Theoretical analysis and simulation show the IMM algorithm is one of the most cost-effective schemes in all generalized pseudo Bayesian algorithms.

IMM估计方法通过在用于输入综合的模型预测概率中引入一个马尔可夫先验转移概率，使得IMM估计不但充分地利用了系统的后验信息，而且合理地利用了系统的先验信息，从而大大地提高了估计的精度和性能。理论分析和仿真结果表明：IMM估计方法是各种广义伪贝叶斯算法中性能最优的一种自适应估计算法。

Describes general methods of point and interval parameter estimation and the small and large sample properties of estimators: method of moments, maximum likelihood, unbiased estimation, Rao-Blackwell and Lehmann-Scheffe theorems, information inequality, asymptotic relative efficiency of estimators.

点估计和区间估计的一般方法，估计量的小样本和大样本性质：矩法，极大似然估计，无偏估计，Rao-Blackwell 和 Lehmann-Scheffe 理论，信息不等式，渐进相对有效估计量。

Maximum likelihood, unbiased estimation, Rao-Blackwell and Lehmann-Scheffe theorems, information inequality, asymptotic relative efficiency of estimators.

点估计和区间估计的一般方法，估计量的小样本和大样本性质：矩法，极大似然估计，无偏估计，Rao-Blackwell 和Lehmann-Scheffe 理论，信息不等式，渐进相对有效估计量。

Limit theorems, asymptotic methods, asymptotic efficiency and efficiency bounds for estimation, maximum likelihood estimation, Bayes methods, asymptotics via derivatives of functionals, sample-based estimates of variability:; robustness; estimation for dependent data, nonparametric estimation and testing.

高级统计推断理论：极限理论，渐进方法，渐进有效和估计的效率，加大似然估计，贝叶斯方法，泛函数的微商渐进性，基于样本的估计：(bootstrap 和 jackknife方法)；稳健性；独立数据的估计，非参数估计和经验。

Limit theorems, asymptotic methods, asymptotic efficiency and efficiency bounds for estimation, maximum likelihood estimation, Bayes methods, asymptotics via derivatives of functionals, sample-based estimates of variability:; robustness; estimation for dependent data, nonparametric

高级统计推断理论：极限理论，渐进方法，渐进有效和估计的效率，加大似然估计，贝叶斯方法，泛函数的微商渐进性，基于样本的估计：（bootstrap 和 jackknife方法）；稳健性；独立数据的估计，非参数估计和经验。

Under some regular conditions, we propose estimating equations for H,β and θ, and computational algorithm for solving the equations. We prove that the estimators are asymptotically normally distributed.

进而我们提出了估计H、β和θ的一套估计方程，给出了求解这些方程的计算机算法，讨论了估计量的极限性质，证明了估计量是渐近正态的，并且其极限协方差阵可以被相合地估计。

The examples indicate that the unkown line parameters can be estimated faster and more accurately with the proposed objective function and SPSO and the solving procedure relies less on the augmented jacobian matrix.

仿真算例表明：修正目标函数和SPSO算法能快速准确地估计线路参数，且可提高估计准确度和降低对量测系统参数可估计性的要求；取连续多次估计的均值，则可减小量测误差导致的参数估计偏差。

An improved derivative estimation algorithm with a low-pass filter that can suppress the influence of high-frequency noise is suggested to remove potential drawbacks of the backward difference that was applied in neural network inversion dynamic compensation method, such as large estimation error in the presence of noise and price increase of hardware realization when high estimation accuracy is required.

针对神经网络逆动态补偿策略中基于向后差分的微分估计算法在噪声环境下存在较大的微分估计误差，且提高估计精度会增加硬件成本等缺陷，提出了改进的微分估计算法，并以加速度传感器为例，设计了动态补偿实验装置，给出了基于改进的微分估计算法的动态补偿算法的实现步骤。

Secondly, 4 methods of azimuth estimation based on vector hydrophone according to forms of nosie and signal are brought forward. The principles of these 4methods are set forth, simulating them by computer. Acoustic intensity averager can effectively suppress isotropic non-coherent interference; DOA based on line-spectra can effectively detect the target radiating line-spectra and estimate its azimuth angle; Bar graph approach can suppress line-spectra coherent interference and distinguish different targets radiating line-spectrum. Weighted bargraph approach can detect target in broadband coherent interference and estimate its azimuth angle.

其次，根据背景干扰和信号样式的不同，提出了单矢量水听器方位估计的4种方法；阐述了这4种方法的原理，并进行了计算机仿真，指出平均声强器方位估计适用于各向同性噪声场背景干扰，讨论了平均声强器的处理增益；目标辐射噪声中有线谱时，线谱方位估计有很好的性能，其中的LOFAR线谱和高频解调得到的DEMON线谱都是目标的特征谱线，可以利用它们进行方位估计，详细讨论了利用特征线谱单个矢量水听器分辨多目标的性能；直方图方位估计能抗强线谱相干干扰；加权直方图方位估计可以抗宽带相干干扰。

As a differential detection is used between the intercarrier, the proposed algorithm overcome the high sensitivity to timing error of conventional algorithm and is robust to multi-path. The effect of sample frequency offset can't be ignored in a continuous OFDM system of a large number of subcarriers. The estimation precision of the conventional algorithm is low and the realization is complex. A joint estimation of sample frequency offset and fine frequency offset based on difference operation is proposed. As averaging use a large number of samples in a maximum likelihood algorithm, the estimation precision is very high. The use of difference operation cancel the effect of carrier frequency offset, and the proposed sampling frequency estimation algorithm is robust to carrier frequency offset.

最后，本文对连续系统中的采样频率和载波频率精同步技术进行较为深入的研究，由于子载波数目大的连续OFDM传输系统中，采样频率偏差不可忽视，以往的采样频率估计方法精度低、实现复杂，因此本文提出了一种基于差分处理的采样频率偏差和精载波频率偏差联合估计算法，利用了最大似然算法进行采样频率偏差的估计，由于采用了差分处理方法，消除了载波频率偏差对采样频率偏差估计的影响，采用的最大似然算法对大量样点进行平均化处理，所以，相对以往算法，提出的联合算法估计精度高，而且采样频率偏差估计对载波频率偏差具有鲁棒性。

Methods: Five patients with parkinsonism or dystonia were assigned to general anesthesia using an modified endotracheal tube.

本实验依照人体实验之相关规定进行，五位患有帕金森氏症或肌张力异常的病人接受神经立体定位手术。

If you can benefit from this book, it is our honour.

如果您能从本书获益，这将是我们的荣幸。