In this paper, we propose a new method for noise robustness base on Gaussian Mixture Model (GMM), and the method we proposed can estimate the noise feature effectively and reduce noise effect by plain fashion, and we can retain the smoothing and continuity from original feature in this way. Compared to the traditional feature transformation method MMSE(Minimum Mean Square Error) which want to find a clean one, the different is that the method we proposed only need to fine noise feature or the margin of noise effect and subtract the noise to achieve more robustness effect than traditional methods. In the experiment method, the test data pass through the trained noise classifier to judge the noise type and SNR, and according to the result of classifier to choose the corresponding transformation model and generate the noise feature by this model, and then we can use different weight linear combination to generate noise feature, and finally apply simple subtraction to achieve noise reduction. In the experiment, we use AURORA 2.0 corpus to estimate noise robustness performance, and using traditional method can achieve 36:8% relative improvement than default, and the our method can achieve 52:5% relative improvement, and compared to the traditional method our method can attain 24:9% relative improvement.

Our national commercial and cargo airlines have great revenue because of rapid developed economics. Especially after the open sky policy. Aviation capacity grows very fast. The convenient of air transportation brings booming economics to the general public; however it also brings destructive environment of airport neighborhood, lower the living quality. Under the aircraft noise for a long term would affect people¡¦s life and work seriously, cause people different level of physical and psychological damages. Citizen views the present monitoring operation of aircraft noise and causes serious environmental pollution issues. Although the government studies airside noise improvement and reward system aggressively, the contestation continues for three years. Every individual airport has conflict with close-by residents frequently. The issue of rationality and fairness of the official aircraft noise prevention and airport reward system is important and meet the exigencies of the moment. This study views the past three years document of aircraft noise prevention and airport reward system, refer to varies information of national and international similar system, analysis and study the reasonable standard level of aircraft noise and related regulations for schools, hospitals residents. Further to offering some strategies and methods to improve the aircraft noise prevention and airport reward system.

A novel method to reduce speckle noise from a digital image is presented. Speckle noise is introduced once a coherent light source is used. In this paper, we use the Empirical Mode Decomposition(EMD) method to remove speckles caused by such kind of coherent illumination. Many filter algorithms, such as Band-pass Filter, Enhanced Frost Filter, Gamma Filter, Enhanced Lee Filter, have been extensively studied to remove the speckle. However, they cannot remove noise effectively. The EMD method is able to analysis noise efficiently. This makes it possible to accurately analyze fringes in the frequency domain and to accurately retrieve the signal.

Of all the forms of radiation known, sound travels through the sea the best, therefore it is widely used in underwater detection and signal transmission. However, the ambient noise in the ocean will affect the quality of these applications, so that understanding of ambient noise is important for improving acoustic capability. Previous literatures show that ambient noise at deep water should decrease with depth, but would it be the same in the shallow? Most studies of ambient noise were made at deep water in the past, the reason is that great depth variation of ambient noise is easily observed. The ASIAEX 2001 South China Sea experiment is on the continental shelf-break, so the depth effect of ambient noise in shallow water can be explored. A typhoon shall affect experimental zone during May 10-14, 2001. From the results of this study, afternoon effect shall affect distribution of temperature and ambient noise under water depth. The other, a typhoon can not only affect seriously ambient noise, but also intercept the shine and variation of sea temperature, and reduce afternoon effect. Another, ambient noise has insignificant change with depth, however it¡¦s is much affected by the variation of temperature, frequency, and occurrence of internal wave.

A general framework for performance optimization of continuous-time OTA-C (Operational Transconductance Amplifier-Capacitor) filters is proposed. Efficient procedures for evaluating nonlinear distortion and noise valid for any filter of arbitrary order are developed based on the matrix description of a general OTA-C filter model . Since these procedures use OTA macromodels, they can be used to obtain the results significantly faster than transistor-level simulation. In the case of transient analysis, the speed-up may be as much as three orders of magnitude without almost no loss of accuracy. This makes it possible to carry out direct numerical optimization of OTA-C filters with respect to important characteristics such as noise performance, THD, IM3, DR or SNR. On the other hand, the general OTA-C filter model allows us to apply matrix transforms that manipulate (rescale) filter element values and/or change topology without changing its transfer function. The above features are a basis to build automated optimization procedures for OTA-C filters. In particular, a systematic optimization procedure using equivalence transformations is proposed. The research also proposes suitable software implementations of the optimization process. The first part of the research proposes a general performance optimization procedure and to verify the process two application type examples are mentioned. An application example of the proposed approach to optimal block sequencing and gain distribution of 8th order cascade Butterworth filter (for two variants of OTA topologies) is given. Secondly the modeling tool is used to select the best suitable topology for a 5th order Bessel Low Pass Filter. Theoretical results are verified by comparing to transistor-level simulation withCADENCE. For the purpose of verification, the filters have also been fabricated in standard 0.5mm CMOS process. The second part of the research proposes a new linearization technique to improve the linearity of an OTA using an Active Error Feedforward technique. Most present day applications require very high linear circuits combined with low noise and low power consumption. An OTA based biquad filter has also been fabricated in 0.35mm CMOS process. The measurement results for the filter and the stand alone OTA have been discussed. The research focuses on these issues.

This thesis addresses the recurrent threshold crossing behavior of long-time correlated noise. The behavior of long-time correlated noise like f / 1 , 5 . 1 / 1 f , and 2 / 1 f can be associated with the behavior of many phenomena in nature, so it is of interest to study the behavior of this noise. Our method of modeling their recurring behavior relies on setting a particular threshold level for a particular level of noise and observing how frequently the noise crosses the threshold level. We also add a periodic drive to the noise which enables it to cross the threshold level easily when it is at peak, and vice versa. This technique provides a model for the changing seasons that occur during every year. We also compare the recurrence behavior of threshold crossings from our computer simulations with theoretical results from the Rice formula. We have related the recurrence of these threshold crossings with the recurrence of natural disasters. Therefore we are providing a model to predict the recurrence of a natural disaster once that disaster has previously occurred. From our results, we conclude that once a natural disaster has occurred, there is a high probability of its recurrence in a short time, and this probability gradually decreases with time.

Measuring the farfield noise levels of full-scale rotor systems is not trivial and can be costly. Researchers prefer to perform small-scale experiments in the laboratory so that they can extrapolate the model scaled results to the larger scale. Typically Inverse Square Law (ISL) is used to extrapolate the sound pressure levels (SPL), obtained from model-scale experiments at relatively small distances to predict noise at much larger distances for larger scale systems. The assumption underlying this extrapolation is that the source itself can be treated as a point sound source. At what distance from a rotor system it can be treated as a point source has never been established. Likewise, many theoretical models of shielding by hard surfaces assume the source to be a point monopole source. If one is interested in shielding the noise of a rotor system by interposing a hard surface between the rotor and the observer, can the rotor system really be considered to be a monopole? If rotating noise sources are under consideration what is the effect of configuration and design parameters? Exploring the validity of point source assumption alluded to above for a rotor for farfield acoustic measurements with and without shielding form the backbone of the present work.

Thesis (doctoral)--Helsinki University of Technology, 1999. / Includes bibliographical references. Also available on the World Wide Web.

published_or_final_version / abstract / toc / Mechanical Engineering / Doctoral / Doctor of Philosophy

No description available.