在算法上

The method of constructing arithmetic of oil system steady model for aero engine has been elaborated in this paper, which is based on the analyses of different parts characteristics that show monotone variation relationship and which introduces the notion "balanced calculation" to replace traditional iteration with vector calculus and various interpolation algorithms.

在分析滑油系统部附件的基础上，根据部附件特性单调变化的特点，提出了"平衡计算"的概念，应用向量运算方法和多种插值算法取代传统的迭代算法来构造航空发动机滑油系统稳态模型算法的思路和方法，并用该算法编制程序对发动机在海平面最大状态下滑油系统性能参数进行了验算。

Then, the relation between vibration and sound, the acoustic holography method, the indeterminacy and identifiability of BSS, the BSS algorithm are discussed in detail. Based on above researches, the dissertation is divided in following four sections. The first section investigates the BSS algorithm suitable to acoustic feature separation. The algorithm of the joint approximate diagonalization of eigen-matrices is proposed. The spectra or the time-frequency distributions of source signals are the interesting features in ABD and they are separated from mixing signals by JADE algorithm. Then, the convolutive mixing model is transformed into a high-dimension instantaneous mixing model, and the deconvolution of source signals is achieved by the joint approximate block diagonalization of eigen-matrices. The proposed algorithm has a global minimum, and it is unsensitive to noise interference.

论文首先概述国内外声学诊断研究进展与盲分离研究进展，给出机器噪声声场与盲分离的数学描述，讨论了声振辐射、声全息、盲分离模型、盲分离可解性、盲分离结果不确定性、分离算法等基本问题，在此基础上，论文的研究工作分为以下四个部分：第一部分研究适用于声学特征分离的盲分离算法，提出基于特征提取的联合近似对角化盲分离算法，该算法以频谱特征或时频特征作为分离目标，从混合信号中分离源信号频谱特征或时频特征，最大限度地保留了与声学特征提取有关的频谱特征或时频特征，采用模型变换把卷积混合模型变换为一个高维瞬时混合模型，通过联合近似分块对角化算法实现源信号频谱特征与时频特征的盲反卷积。

On this basis, exploring the method and feasibility to extract terrestrial parameters from hyperspectral and multiangular data. On the inverse problem, we analyzed the error source of model-based inversion, and pointed out the effect of a priori information, merit function, inverse algorithm and strategy on the inverse result.

本文在综合考虑各种算法的优缺点基础上，将基于实编码的遗传算法引入遥感模型反演中，并通过与确定性搜索算法相比较，肯定了遗传算法在解决非线性反演问题中的作用。

In order to study these algorithms systematically and deeply, they are reviewed in this paper based on c-means algorithm, from metrics, entropy, and constraints on membership function or cluster centers. Moreover, the advantages and disadvantages of the typical fuzzy partitional algorithms are discussed.

为了能更为系统和深入地了解这些聚类算法及其性质，本文从改变度量方式、改变约束条件、在目标函数中引入熵以及考虑对聚类中心进行约束等几个方面，对在C-均值算法的基础上得到的基于划分的模糊聚类算法作了综述和评价，对各典型算法的优缺点进行了实验比较分析。

Simulation results show that the proposed method has good performance in time delay predictation of networked control systems, and the plant output can trace desired output effectively.(2)By combining node buffer queuing management and adaptive predictive control, a novel approach on predictive control of NCS is proposed.

在本方法中，使用了直接辨识控制器参数的自适应预测控制算法，该算法不需要在线求解Diophantine方程，减少了计算时间，较好地保证了NCS的实时性；在此基础上，又进一步提出了一种改进型的自适应预测控制算法，不仅避免了在线求解Diophantine方程，并且也不需要进行矩阵求逆运算，使算法的实时性大大提高。

Some optimized schemes can instruct nature gas pipeline network systems routine dispatching and management to increase nature gas pipeline network system economic benefit and fully utilize pipeline network system transportability.

通过求解天然气管网系统稳态运行优化模型，验证了这三种算法的可行性；(3)在研究混合遗传退火算法混合策略的基础上，形成了两种混合算法，混合退火复合形算法和混合遗传复合形算法。

On the one hand, de noising of instantaneous parameters is optimized by a wavelet filter, which can improve the recognition ability under the low SNR; on the other hand, the existing three feature parameters σ,σ(subscript ∞) and σ are used, and three new parameters Ra, Rp and Rf are proposed based on the existing feature parameters, which can simplify the process and improve the recognition ability.

该算法一方面提出采用小波滤波器对信号的瞬时信息进行消噪处理，提高了调制识别算法在低信噪比下的识别能力；另一方面利用已有算法的3个特征参数σ、σ、和σ以及在已有算法基础上改进的3个新的特征参数Ra、Rp和Rf，简化了识别过程，改善了低信噪比下的识别效果。

Then a BGPEOpt(BGP Egress Optimization Algorithm) algorithm is proposed to reduce the execution time of the typical algorithm by leveraging the .characteristics of cut link in the topology graph to reduce the frequency of recomputing traffic demand. Further more, mediate results are used in the iterative process to reduce the time of computing the IGP shortest distances from a source node to other nodes. Experimental results indicate that the running time of BGPEOpt reduced 7%~30% with the similar link utilization comparing to that of typical algorithms.

在此基础上，提出了静态阶段的BGP出口选择优化算法BGPEOpt(BGP Egress Optimization Algorithm)，BGPEOpt算法利用网络拓扑结构中割边的特殊性，减少流量矩阵重新计算的频率；同时利用算法在迭代过程中的中间结果，减少重新计算从一个源节点到其它节点最短路径的时间，从而提高了算法的时间性能。

In this article, we firstly review the history that Wavelet developed, after that, we discuss filter algorithm including FFT algorithm, Fast-running algorithm and Distributed algorithm.

在本文中，首先回顾了小波发展的历史，然后讨论了滤波算法，较详细了分析了快速傅立叶算法、快行滤波算法和分布式算法，在此基础上，简要的讨论了小波变换的单层分解。

When a node in hypercubes consists of a main procesor and a communication coprocessor, link fault model is more adequate for routing in injured hypercubes.

本文在研究了各种有关选路算法的基础上，提出了三种具有不同特点的容错选路算法：局部启发式容错选路算法[128，129]，全局引导式容错选路算法和双向搜索式容错选路算法[130]。

The split between the two groups can hardly be papered over.

这两个团体间的分歧难以掩饰。

This approach not only encourages a greater number of responses, but minimizes the likelihood of stale groupthink.

这种做法不仅鼓励了更多的反应，而且减少跟风的可能性。