信号与系统 第五章课件
- 格式:pps
- 大小:7.24 MB
- 文档页数:66
《信号与系统教案》PPT课件第一章:信号与系统概述1.1 信号的概念与分类信号的定义信号的分类:连续信号、离散信号、随机信号等1.2 系统的概念与分类系统的定义系统的分类:线性系统、非线性系统、时不变系统、时变系统等1.3 信号与系统的研究方法解析法数值法图形法第二章:连续信号及其运算2.1 连续信号的基本性质连续信号的定义与图形连续信号的周期性、奇偶性、能量与功率等性质2.2 连续信号的运算叠加运算卷积运算2.3 连续信号的变换傅里叶变换拉普拉斯变换Z变换第三章:离散信号及其运算3.1 离散信号的基本性质离散信号的定义与图形离散信号的周期性、奇偶性、能量与功率等性质3.2 离散信号的运算叠加运算卷积运算3.3 离散信号的变换离散时间傅里叶变换离散时间拉普拉斯变换离散时间Z变换第四章:线性时不变系统的特性4.1 线性时不变系统的定义与性质线性时不变系统的定义线性时不变系统的性质:叠加原理、时不变性等4.2 线性时不变系统的转移函数转移函数的定义与性质转移函数的绘制方法4.3 线性时不变系统的响应输入信号与系统响应的关系系统的稳态响应与瞬态响应第五章:信号与系统的应用5.1 信号处理的应用信号滤波信号采样与恢复5.2 系统控制的应用线性系统的控制原理PID控制器的设计与应用5.3 通信系统的应用模拟通信系统数字通信系统第六章:傅里叶级数6.1 傅里叶级数的概念傅里叶级数的定义傅里叶级数的使用条件6.2 傅里叶级数的展开周期信号的傅里叶级数展开非周期信号的傅里叶级数展开6.3 傅里叶级数的应用周期信号分析信号的频谱分析第七章:傅里叶变换7.1 傅里叶变换的概念傅里叶变换的定义傅里叶变换的性质7.2 傅里叶变换的运算傅里叶变换的计算方法傅里叶变换的逆变换7.3 傅里叶变换的应用信号分析与处理图像处理第八章:拉普拉斯变换8.1 拉普拉斯变换的概念拉普拉斯变换的定义拉普拉斯变换的性质8.2 拉普拉斯变换的运算拉普拉斯变换的计算方法拉普拉斯变换的逆变换8.3 拉普拉斯变换的应用控制系统分析信号的滤波与去噪第九章:Z变换9.1 Z变换的概念Z变换的定义Z变换的性质9.2 Z变换的运算Z变换的计算方法Z变换的逆变换9.3 Z变换的应用数字信号处理通信系统分析第十章:现代信号处理技术10.1 数字信号处理的概念数字信号处理的定义数字信号处理的特点10.2 现代信号处理技术快速傅里叶变换(FFT)数字滤波器设计数字信号处理的应用第十一章:随机信号与噪声11.1 随机信号的概念随机信号的定义随机信号的分类:窄带信号、宽带信号等11.2 随机信号的统计特性均值、方差、相关函数等随机信号的功率谱11.3 噪声的概念与分类噪声的定义噪声的分类:白噪声、带噪声等第十二章:线性系统理论12.1 线性系统的状态空间描述状态空间模型的定义与组成线性系统的性质与方程12.2 线性系统的传递函数传递函数的定义与性质传递函数的绘制方法12.3 线性系统的稳定性分析系统稳定性的定义与条件劳斯-赫尔维茨准则第十三章:非线性系统13.1 非线性系统的基本概念非线性系统的定义与特点非线性系统的分类13.2 非线性系统的数学模型非线性微分方程与差分方程非线性系统的相平面分析13.3 非线性系统的分析方法描述法映射法相平面法第十四章:现代控制系统14.1 现代控制系统的基本概念现代控制系统的定义与特点现代控制系统的设计方法14.2 模糊控制系统模糊控制系统的定义与原理模糊控制系统的结构与设计14.3 神经网络控制系统神经网络控制系统的定义与原理神经网络控制系统的结构与设计第十五章:信号与系统的实验与实践15.1 信号与系统的实验设备与原理信号发生器与接收器信号处理实验装置15.2 信号与系统的实验项目信号的采样与恢复实验信号滤波实验信号分析与处理实验15.3 信号与系统的实践应用通信系统的设计与实现控制系统的设计与实现重点和难点解析信号与系统的基本概念:理解信号与系统的定义、分类及其研究方法。
Ch5 The Discrete-Time Fourier Transform 第5章离散时间傅立叶变换Abbreviations(缩写):1. CFS :The Continuous-Time Fourier Series ——连续时间傅立叶级数2. DFS :The Discrete-Time Fourier Series ——离散时间傅立叶级数3. CTFT :The Continuous-Time Fourier Transform ——连续时间傅立叶变换4. DTFT :The Discrete-Time Fourier Transform ——离散时间傅立叶变换Focus on itin thisMain content1.Development of the Discrete-Time Fourier Transform (离散时间傅立叶变换的导出);2.Basic Fourier Transform Pairs (常用信号的离散时间傅立叶变换对);3. The Fourier Transform for Periodic Signals (离散时间周期信号的傅立叶变换);4. Properties of the Discrete-Time Fourier Transform (傅立叶变换的性质);5. The frequency response and frequency-domain methods for discrete-time signals and systems (离散系统的频率响应与频域分析方法);5.0 IntroductionAnalytical objects : aperiodic discrete-time signals and systemsAnalytical methods: (similar to CTFT)1)An aperiodic d-t signal can be viewed a periodic d-t signals with an infinite period.2)As the period becomes infinite, the discrete-time Fourier series represen-tation becomes the discrete-time Fourier transform .3)Using the property of the Fourier transform to examine frequency-domain analysis of signals and LTI systems.note that the similarities and differences between continuous-time and discrete-time Fourier transform analysis.5.1 Representation of Aperiodic Signals: The Discrete-Time Fourier Transform(p359)(非周期信号的表达:离散傅里叶变换)5.1.1Development of the Discrete-Time Fourier Transform(离散傅里叶变换的导出)Consider a discrete-time periodic signal:N→∞periodic aperiodic[][]221,j kn j kn N N k k k N n N x n a e a x n e N ππ-=<>=<>==∑∑for a d-t periodic signal , we have the discrete-time Fourier series pair:[]x n [][]22/2/2N j kn j kn N N k n N n Na x n e x n e ππ+∞--=-=-∞==∑∑so discrete-time Fourier transform[]j j n n X e x n e ωω∞-=-∞=∑() , 2N k N πω→∞→ as lim j k N Na X e ω→∞()Define:so []00000012(),1()2jk jk n k N jk jk n k N x n X e e N N X e e ωωωωπωωπ=<>=<>=⋅==⋅⋅∑∑0211()()jk j k k N a X e X e N N ωωπω===Compare with ,k a j X e ω()we have []212j j n x n X(e )e d ωωπωπ∴=⎰N →∞asan aperiodic sequence can be thought of as a linear combination of complex exponentials infinitesimally close infrequency and with amplitudes 12j X(e )d ωωπ[]212j j n x n X(e )e d ωωπωπ=⎰[]j j nX(e )x n e ωω∞--∞=∑结论:discrete-time Fourier transform pairDifferences between the c-t and d-t Fourier transform :2) the finite interval of integration in the synthesis equation.1) periodicity of the discrete-timetransform j X(e )ω(2)()()j j X eX e ωπω+=In discrete-time ,Low frequencies are the values of near evenmultiple of ;high frequencies are the values of near oddmultiples of .ωπωπLow-frequency signalhigh-frequency signal5.1.2 examples of discrete-time Fourier transform (p362)(离散傅里叶变换的例子)[]()11nj nj nj j n n X(e)a u n eaeaeωωωω∞∞---=-∞====-∑∑Where is a complex function j X(e )ω11j a sin X(e)tga cos ωωω-=--Example 5.1(p362)[][]1nx n a u n ,a =<2112j X(e )a acos ωω=+-Magnitude:Phase:All of values are of theDecaying same signLow-frequencysignal<<01ahigh-frequency signal10a -<<alternate in valueDecayingExample 5.2 (p364)[]1nx n a ,a =<[][][]1nnx n a u n a u n -=--+1122111112j n j nn j nn n nj nn j nn n j j j X (e)aea ea ea eae a ae ae a a cos ωωωωωωωωω-∞---=-∞=∞∞-==-=+=+-=+=--+-∑∑∑∑01a <<Real and even sequenceReal and even functionLow-frequency signalExample 5.3(p365)[]1110n N ,xn ,n N ≤⎧=⎨>⎩1112122N j j nn N sin(N )X (e)esinωωωω-=-+==∑12N Real and even sequenceReal and even function1211k sin k(N )N a ,N sin k Nππ+=①Compare with the corresponding periodic square wave signal21j k k Na X (e )N ωπω==so 12122j sin(N )X(e )sin ωωω+=②Compare with the corresponding c-t aperiodic square wave signal1112T sin T X(j )T ωωω=12122j sin(N )X(e )sin ωωω+=Example 5.4(p367)[][]x n n δ=[]n δ0n1[]()1j j nn X e x n eωω∞-=-∞==∑()j X e ω1π-πω5.1.3Convergence Issues associated with the Discrete-Time Fourier Transform (p366)(离散傅里叶变换的收敛问题)Question:[]j j nX(e )x n e ωω∞--∞=∑①What is the conditions on thatguarantee the convergence of thissum []x n Answer 1:Condition:[]n x n +∞=-∞<∞∑( i.e. is absolutely summable)[]x nAnswer 2:Condition:[]2n x n +∞=-∞<∞∑( i.e. the sequence has finite energy)The two conditions above can guarantee the convergence of []j j nX(e )x n eωω∞--∞=∑②Are there convergence issuesassociated with ?[]212j j n x n X(e )e d ωωπωπ=⎰NO!Because the integral in this equation is over a finite interval of integration.We approximate by[]x n[]()12Wj j nWˆx n X e e dωωωπ+-=⎰[]nδ5.2 The Fourier Transform For Periodic Signals(p367)(周期信号的离散傅里叶变换)consider the Fourier transform of the sequence0[]j nx n eω=In c-t time, we saw ()002j teωπδωω↔-(Note the d-t Fourier transform must be periodic in with )ω2π0022j nk ek ωπδωωπ∞=-∞↔--∑()Therefore, we expect[]002,jk nk k N x n a eNωπω=<>==∑[]0()2(2)j kk N l x n X e a k l ωπδωωπ∞=<>=-∞↔=--∑∑02021()(-)2j nj j nj nX e ed ed eπωωωωπωδωωωπ==⎰⎰To check the validity of the expression above,consider an arbitrary periodic sequence []x n so0022j nk ek ωπδωωπ∞=-∞↔--∑()()22()j k k X e a k N ωππδω∞=-∞=-∑Example 5.5 :(P371)00011[]cos 22j n j n x n n e e ωωω-==+025πω=with 22()(2)(2)55j k X e k k ωπππδωπδωπ∞=-∞⎡⎤←−→=--++-⎢⎥⎣⎦∑F22()()()55j X e ωπππδωπδωπωπ=-++-≤<That isExample 5.6 :(P371)[][]k x n n kN δ∞=-∞=-∑periodic impulse train [][]001111N jk njk nk n N n a x n en eNNNωωδ---=<>====∑∑22()j k X e k NN ωππδω∞=-∞⎛⎫=- ⎪⎝⎭∑5.3 Properties of The Discrete-Time Fourier Transform(P372)(离散时间傅立叶变换的性质)Adopting notation below to indicate the pairing of a signal and its transform[]()j x n X eω←−→F5.3.1Periodicity (周期性)(2)()()j j X eX e ωπω+=5.3.2Linearity (线性)5.3.3Time Shifting and Frequency Shifting(时移和频移)[][]1212()()j j ax n bx n aX e bX e ωω+↔+[]()j x n X e ω←−→Fthen[]()j x n X e ω↔00()j n j x n n X e e ωω--←−→⎡⎤⎣⎦F Time Shifting []00()()j nj x n eX eωωω-←−→FFrequencyShiftingSolution:Determine the relationship of the impulseresponses of an ideal lowpass filter and an ideal highpass filter .()j hpH e ω()j lp H e ωExample 5.7(p374)Shifting by ()()()j j h p l p H e H eωωπ-=()j lp H e ωπBy frequency shifting property,[][]()[]1nj nh p l p l p h n eh n h n π==-5.3.4Conjugation and Conjugate Symmetry (共轭及共轭对称)[]()j x n X e ω←−→F[]**()j x n X eω-←−→F*()()j j X e X eωω-=When is real[]x n Re ()Re ()Im ()Im ()j j j j X e X e X e X e ωωωω--⎧⎡⎤⎡⎤=⎪⎣⎦⎣⎦⎨⎡⎤⎡⎤=-⎪⎣⎦⎣⎦⎩()()()()j j j j X e X e X e X e ωωωω--⎧=⎪∴⎨=-⎪⎩{}{}[]()j x n X e ω←−→FEv Re {}{}[]()j x n j X e ω←−→FOd Im5.3.5Differencing and Accumulation (差分与求和)[]1(2)1j k u n k e ωπδωπ∞-=-∞∴←−→+--∑F[][]nk u n k δ=-∞=∑[]1n δ←−→F[][][]1(1)()1()(2)1j j nj j k k x n x n eX e x k X e k e ωωωωπδωπ-∞-=-∞=-∞--←−→-⎡⎤←−→+-⎢⎥-⎣⎦∑∑FFExample 5.8 :(P376)Determine []u n ←−→F?[]()j x n X eω--←−→F[]()j x n X e ω←−→F5.3.6Time Reversal (时域反转)If is real and even ,then[]x n ()()()j j j X e X eX e ωωω-*==i.e. is real and even.()j X e ωIf is real and odd ,then[]x n i.e. is purely imaginary and odd.()()()j j j X e X e X e ωωω-*=-=-()j X e ω5.3.7Time Expansion (时域扩展)()[/],[]0,k x n k if n is a multiple of kx n if n is not a multiple of k ⎧=⎨⎩[][]()j j nj rkk kkn r X e x n ex rk eωωω∞∞--=-∞=-∞==∑∑[]()j rkjk r x r eX eωω∞-=-∞==∑[]()jk k x n X eω∴↔inverse relationship[]x n Example 5.9(p379)Determine of ()j X eω(2)(2)[][]2[1]x n y n y n =+-Solution:2sin(5/2)()sin(/2)j j Y e eωωωω-=4(2)sin(5)[]sin()F j y n eωωω-←−→5(2)sin(5)2[1]2sin()F j y n eωωω--←−→()4()sin(5)12sin()j j j X e eeωωωωω--⎛⎫=+ ⎪⎝⎭power-density spectrumEnergy-density spectrum5.3.8Differentiation in Frequency (频域微分)[]()j dX e jnx n d ωω-↔5.3.9Parseval’s Relation (帕斯瓦尔定理)[]22212()j n x n X e d ωπωπ∞=-∞=∑⎰[]221kn N k N x n a N=<>=<>=∑∑Compare with DFS①The convolution property maps (映射)the convolution of two signals to the simple algebraic operation of multiplying their Fourier transforms.5.4 The Convolution Property (P382)(卷积性质)[][][]*()()()j j j y n x n h n Y e X e H e ωωω==if thenImpulse response of LTI systemInput of LIT systemoutput of LITsystemFrequency response of LTI system()j H eω✓②The frequency response captures the change in complex amplitude of the Fourier transform of the input at each frequency ω✓③The convolution property is the basis of the frequency domain analysis of signals and systems.Consider an LTI system with 0[][]h n n n δ=-Analyze that how the system impacts on the input in frequency domain?Solution:0()[]j n j j nn H e n n eeωωωδ+∞--=-∞=-=∑The frequency response is()()j n j j Y e eX e ωωω-=Therefore0[][]y n x n n =-Using time-shifting property, A pure time shift system()j n j H e eωω-=Magnitude = 1 at all frequenciesphase =that is linear with frequency .0n ω-Consider the d-t ideal lowpass filter[]?h n =Solution:[]()()1122sin ccj j nj j nc h n H e ed He e d nnωπωωωωπωωωππωπ++--===⎰⎰Not causaloscillatory5. 5 The Multiplication Property (P388)(相乘性质)[][][]12()122121()()()21()()2j j j j j y n x n x n Y e X e X e d X e X e ωθωθπωωθππ-=⋅==⊗⎰ifthen periodic convolution5. 6 TABLES OF FOURIER TRANSFORM PROPERTIES AND BASIC FOURIER TRANSFORM PAIRS(P390)(傅立叶变换的性质及基本变换对列表)5. 7 duality(P390)(对偶)5.7 对偶性(Duality )∑∑>=<->=<==N n nNjk k N k n Njk ken x Na ea n x ππ22)(1,)(由于本身也是以N 为周期的序列,当然也可以将其展开成DFS 形式。