|
| |
|
|
Echo suppressor
Kratochvíl, Pavel ; Smékal, Zdeněk (referee) ; Sysel, Petr (advisor)
In the communication networks a common problem is a returned copy of an original signal that goes back through the network to the sender, disturbing the communication. This problem is attacked using Echo Cancellors that should comply with the specifications of ITU-T organization. Algorithms suitable for cancellors should be tested under specified conditions, then a decision should be taken regarding their deployment. This work deals with basic Least Mean Squares and Normalized Least Mean Squares algorithms.
|
|
|
Adaptive Noise Cancellation in Speech Application
Vomáčka, Jan ; Průša, Zdeněk (referee) ; Sysel, Petr (advisor)
This semestral project deal with methods adaptive noise rejection in acoustic recordings, above all speech signals. Work deal with proposal of suitable system for effective remove noise from acoustic recordings. In theoretic parts are out of print principles of adaptive filters and method of repression noise in acoustic recordings. In practical parts come out proposal and system simulation for adaptive noise rejection in speech.
|
|
| |
|
|
Digital signal processing in real time
Zamazal, Zdeněk ; Mačák, Jaromír (referee) ; Rášo, Ondřej (advisor)
This work deals with digital signal processing in the field of adaptive filtering. Fundamental basics of adaptive filtering are described and primary aim is to create executable laboratory examples, using adaptive filtering, in LabView programming language. These laboratory examples are intended to be used by students fo studying and during laboratory lessons. Objective is to connect the examples with external devices, such as microphone. A microphone is used as an user's speech input acquiring interface. In the thesis is depicted Wiener's filter and problem of adaptive filtering is discussed. Contemporary adaptive algorithms are described and their applications as well. Most mentioned is the LMS algorithm and it's forms. Laboratory examples use following concepts: Adaptive Echo Cancellation, Active Noise Control and System Identification. Each of these examples is solely executable (need for LabView or Run-time engine), consisting also of theory with diagrams. Examples therefore are usable even without manual.
|
|
|
Analysis, Implementation and Utilization of the Vold-Kalman Filter for Non-Stationary Signals
Čala, Martin ; Tůma, Jiří (referee) ; Žiaran,, Stanislav (referee) ; Beneš, Petr (advisor)
The doctoral thesis focuses on a Vold-Kalman filter (VKF). Theoretical part describes properties of VKF and other order tracking methods, namely computed order tracking (COT) and Gabor order tracking (GOT). It also characterizes requirements for rotational speed measurements as one of the key elements for correct functionality of VKF. Practical part depicts own filter implementation and its properties. Main stress is put on computational efficiency, that is in result better than in available codes. Thesis also points out possible issues with numerical instabilities within calculation caused by limited dynamic range of double data type. This is solved by restricting the inputs to prevent the instabilities. Restriction is applied also to cases where the result is numerically correct but unusable. Following part extends the comparison with methods STFT, COT and GOT, where benefits of VKF for nonstationary conditions are shown. The last section shows given information used on simulated signals. This is then applied to show mentioned techniques on experimental data, for instance from turbo engine or electric motor, where the ability of VKF in checking the accordance between speed profile and vibration data is illustrated.
|
|
|
Control system for electrodynamic vibration exciter
Čala, Martin ; Beneš, Petr (referee) ; Havránek, Zdeněk (advisor)
Master's thesis deals with implementation of a control system for an electrodynamic exciter. Patent and literature research with focus on applicable control methods is performed. Then, description of the methods applicable in the system identification, as well as types of the signals used in the vibration testing are discussed. The practical section is dedicated to the analysis of available hardware resources, design, assembly and verification of a control system for an electrodynamic exciter equipped with piezoelectric accelerometer and components of National Instruments.
|
|
|
Implementation of the channel equalization algorithms used in FMT modulation
Krejča, Libor ; Krajsa, Ondřej (referee) ; Šilhavý, Pavel (advisor)
The objective of Diploma thesis is design of analysis tool for equalizers used in FMT modulation. The model of transmission channel was designed for simulations with FMT. The transmission path is modeled by test loop, which corresponds to DSL line. For this reason, some principles of DSL technology is described in the thesis. The principles of multicarrier modulation are introduced in first part. The multicarrier modulation with filter bank (FMT) is described in detail.The different methods of design the fiter bank are given and compared. Channel equalization are introduced in second part. The attention was focused on minimum mean square error filtering (MMSE). Decision feedback channel equalizer (DFE) is extended from linear MMSE equalizer. DFE equalizers were programmed in analysis tool. For computation of equalizer coeficients was used also equalizers based on adaptive algorithms and MMSE. The last part describes the results of DFE equalizers used in communication system with FMT modulation. Analysis tool was programmed in MATLAB with a graphical user interface. It allows to show mean square error, signal-to-noise ratio and transmission speed dependence on delay between original and distorted signal. Signal-to-noise ration is displayed also in individual subchannels and users can display mean square error dependence on different orders of DFE filters.
|
|
|
Filtering EEG signals
Pohludka, Aleš ; Ronzhina, Marina (referee) ; Potočňák, Tomáš (advisor)
This work describes basics of electroencephalography, measuring methods of electroencephalographic signals, their processing and especially their filtration. The most basic normalised international system for measuring EEG from the scalp is described. The end of theoretical part is dedicated to the most frequent artifacts. The practical part deals with testing two methods based on adaptive filtration of reference EOG signals using the LMS algorithm. Further in the paper are presented tests of before mentioned methods and their verification in practice on gathered signals from volunteers. At the conclusion there is an evaluation of the methods used.
|
|
|
Analysis, Implementation and Utilization of the Vold-Kalman Filter for Non-Stationary Signals
Čala, Martin ; Tůma, Jiří (referee) ; Žiaran,, Stanislav (referee) ; Beneš, Petr (advisor)
The doctoral thesis focuses on a Vold-Kalman filter (VKF). Theoretical part describes properties of VKF and other order tracking methods, namely computed order tracking (COT) and Gabor order tracking (GOT). It also characterizes requirements for rotational speed measurements as one of the key elements for correct functionality of VKF. Practical part depicts own filter implementation and its properties. Main stress is put on computational efficiency, that is in result better than in available codes. Thesis also points out possible issues with numerical instabilities within calculation caused by limited dynamic range of double data type. This is solved by restricting the inputs to prevent the instabilities. Restriction is applied also to cases where the result is numerically correct but unusable. Following part extends the comparison with methods STFT, COT and GOT, where benefits of VKF for nonstationary conditions are shown. The last section shows given information used on simulated signals. This is then applied to show mentioned techniques on experimental data, for instance from turbo engine or electric motor, where the ability of VKF in checking the accordance between speed profile and vibration data is illustrated.
|
guest ::
