C#access学籍管理系统

纯前端生成二维码（并将Html转换成图片） 下载到本地,需要用到的四个JS文件jquery.qrcode.min.js html2canvas.min.js canvas2image.js base64.js

找工作的 同学没 这是数字IC的经典笔试题，好好看看很有帮助。自己准备的，感觉很全

FAM和SSCA算法的matlab源程序-detection and identification of signalFAM和SSCA算法的matlab源程序-detection and identification of signalDISCLAIMER NOTICEMTHIS DOCUMENT IS BESTQUALITY AVAILABLE. THECOPY FURNISHED TO DTICCONTAINED A SIGNIFICANTNUMBER OF PAGES WHICH DONOT REPRODUCE LEGIBLY.ilApproved for public release; distribution is unlimitedDETECTION AND IDENTIFICATION OF CYCLOSTATIONARY SIGNALSEvandro luiz da costaLieutenant Commander, brazilian NavyB.S., Instituto Militar de Engenharia, 1980Submitted in partial fulfillmentof the requirements for the degree ofMASTER OF SCIENCE IN ELECTRICAL ENGINEERINGANDMASTER OF SCIENCE IN ENGINEERING ACOUSTICSfrom theNAVAL POSTGRADUATE SCHOOLMarch 1996Author.Evandrodk da costaapproved byRQ求Ralph Hippenstiel, Thesis Co-AdvisorRoberto Cristi, Thesis Ca-AdvisoiHerschel. Loomis, Jr, hairmanDepartment of Electrical and Computer EngineeringArthony A. Atchley / ChairmanEngineering Acoustics Academic CommitteeABSTRACTPropeller noise can be modeled as an amplitude modulated(AM) signalCyclic Spectral Analysis has been used successfully to detect the presence ofanalog and digitally modulated signals in communication systems. It can also identithe type of modulation. Programs for Signal Processing based on compiledlanguages such as FORTRAN or C are not user friendly, and MATLAB basedprograms have become the de facto language and tools for signal processingengineers worldwideThis thesis describes the implementation in mAtlab of two fast methods ofcomputing the Spectral Correlation Density(SCD)Function estimate, the FFTAccumulation Method (FAM)and the Strip Spectral Correlation Algorithm( SSCA),toperform Cyclic Analysis. Both methods are based on the Fast Fourier transformFFT)algorithm. The results are presented and areas of possible enhancement forpropeller noise detection and identification are discussedTABLE OF CONTENTSINTRODUCTIONA MOTIVATION,P,中“····*s···:···:B BACKGROUNDC THESIS GOALSIL NOISE IN THE OCEAN··+A TYPES OF UNDERWATER NOISE25561. Ambient Noise番申2. Self noise3. Radiated noise8B RADIATED NOISE FROM SHIPS, SUBMARINES AND TORPEDOES.......8C PROPELLER NOISE10lI CYCLOSTATIONARY PROCESSING15A CYCLOSTATIONARIT15B THE CYCLIC AUTOCORRELATION FUNCTION (ACF)17C THE SPECTRAL CORRELATION DENSITY FUNCTION (SCD)18N. ESTIMATION OF THE SPECTRAL CORRELATION DENSITY FUNCTION23A FFT ACCUMULATION METHOD(FAM)25B STRIP SPECTRAL CORRELATION ALGORITHM(SSCA■D28V. EXPERIMENTAL RESULTSA. ANALOG-MODULATED SIGNALS311. Amplitude Modulated(AM)Signal===2-312. Pulse-Amplitude Modulated(PAM) Signal58B DIGITAL-MODULATED SIGNALS1. Amplitude Shift Keying(ASK) Signal中中非昏号即即号自唱即自曲音非带卡.最662. Binary-Phase Shift Keying(BPSK) Signal6了VI CONCLUSIONS81A SUMMARY81B SUGGESTIONS82APPENDIX A-CALCULATION OF THE SCD FUNCTION OF AN AMPLITUDE-MODULATED SIGNAL83APPENDIX B-FUNCTION AUTOFAM95APPENDIX C-FUNCTION AUTOSSCA99APPENDIXD-FUNCTION CROSSFAM103APPENDIX E-FUNCTION CROSSSSCA109APPENDIX F-PLOTTING ROUTINES113LIST OF REFERENCES115INITIAL DISTRIBUTION LIST看音117INTRODUCTIONA. MOTIVATIONPropeller related acoustic signatures typically exhibit modulationcharacteristics. These modulation characteristics originate from the cavitationprocess that takes place in the water due to the cyclic movement of the propellerThe cavitation process is basically the collapse of air and vapor bubblesdue to variations in the static pressure. These variations in static pressure are aconsequence of the passage of the propeller blades through the water. Thismovement, cyclic in nature, causes amplitude modulation in the static pressureand as a consequence an amplitude-modulated(AM) signal can be detected in areceiverCyclostationary processing techniques have been used to detect andlentify analog and digital communication signals very successfully. Thesetechniques have the advantage of using a more realistic model for the signalthan the stationary model used in most of the more conventional signalprocessing techniquesB BACKGROUNDThe basic elements of cyclic spectral analysis are the time-variant cyclicperiodogram and the time-variant cyclic correlogram. These two functions form aFourier transform pair. This fact is known as the cyclic Wiener relation or thecyclic Wiener-Khinchin relation [Ref. 1: p. 49.1

线性神经网络，BP神经网络，Hopfield神经网格，Elman神经网络，RBF神经网络；在模型应用模块中实现了六种实际应用：RBF网络的船用柴油机故障诊断，BP网络的齿轮箱故障诊断，SOM网络的回热系统故障诊断，BP网络的设备状态分类器，SOM网络的人口比例样本分类，SOM网络的土壤性状样本分类

python项目-face++人脸识别考勤机-python_GUI-automatic_weather-face_gensui-python_game-python_LAN-python_multithreading-python_mysql-python_opencv_tracking-python_pachong

误差分析计算公式及matlab代码实现（均方误差MSE，平均绝对误差MAE，平均绝对百分比误差MAPE，均方百分比误差MSPE，均方根误差RMSE，残差平方和SSE）

使用OpenCV中的calibrateCamera函数进行张正友相机标定，得到相机内参矩阵。

【实例简介】基于FPGA的四层六层PCB板的设计，焊盘过孔尺寸等布线规则，信号完整性分析、仿真以及电路板的设计策略和布线策略。

设一个时隙 Aloha 系统的时隙长度为 1，所有节点的数据包均等长且等于时隙长度。网络中的节点数为 m，各节点数据包以泊松过程到达。 1 假定每个节点的数据包到达强度均为 λ /m，在不同的 λ 下，仿真时隙Aloha 数据包传送的成功概率，绘制呼入强度和成功概率的曲线，和理论结果进行对照。 仿真思路： 1） 生成一个二项分布列来模拟数据包的到达过程 2） 因为数据包以泊松过程到达，所以二项分布的 P 定为（1- m eλ− ） 3） 对生成的数列求和，只有当其和恰等于 1 即有且仅有一个数据包到达时，才可以成功发送，这时成功个数计数+1 4） 2.选取合理的引,,qa,m,采用延时的下界,仿真时隙Aoha系统数据传输过程,统计在不冋同η下,到达率及离开率,绘制它们随n的分布情况,和理论值进行对照qn:等待重传的节点在每一时隙内重传数据包的概率qa:每个发送节点有新数据包到达的概率m:系统内总的节点数n:每个时隙开始时等待重传的节点数仿真思路:1)用二项分布模拟数据包的到达及发送过程2)生成两个数列:一个表示等待重传的节点以q,重传的情况;一个表示新到达的数据包情况因为题日说明采用延时的下界,即不缓冲,每个节点最多容纳一个数据包,有包则扔。所以第一个数列前n项令为1,后一个前n项令为0,之后两个数列可以进行简单加和3)发送成功率:对两个数列相加之后求和,如果sum等于1,说明此时隙内到达和发送的总数为1,只有在这种情况下发送才有可能成功,计数加1到达率:在每N次实验中,对“表示到达的数列”求和,统计4)对n做循环以表示到达率和离开率随n的变化情况;每个n下进行N次实验,数理统计3/8仿真结果0.40.35*0.30.250.20.150.10.0550607080901001/曲线为理论曲线:Ps=G exp(-G)and G=(m-n)a+n gr2/仿真值基木与理论曲线吻合在仿真的过程中,合理选取个参数值对能否得到埋想的曲线起了重要的作用下图分别为qr=0.02,0.05,0.08s时的曲线。可以看到,随着qr的增加,曲线向左移,导致第二个交叉点也左移,这个时候重传的延时将会减小。反之,曲线右移。当q,增加到一定程度的时候,系统只有一个稳定点了。4/840.350.30.20.1501020304050607080901c03仿真时隙Aoha系统下的伪贝叶斯算法,通过仿真结果眼正在n的估计误差较大的情况下的收敛特性及到达率小于1/e下的稳定性。仿真思路:1、伪贝叶斯算法的主要思路是对新数据包和积压节点等同对待:当有新数据包到达的时候,暂不发送,下一时刻与以前的积压节点一起以4r发送。所以修改2中的仿真模型:1)依旧是一列表到达,一列表上一时隙的积压节点2)对两列加和,统计其中为1的个数,设为d3)以qr为概率,d为长度,生成又一个二项分布数列 depart,表示发送的情况4)对depa求和,如果 depart的和为一,说明恰发送成功,n(k+1)=d-1,否则n(k+1)=n(k)5)循环,进行数理统计2、仿真收敛特性和稳定特性哩论值:根据给岀的伪贝叶斯算法的具体步骤,由给出的n(k),不断模拟生成n(k+1)5/8仿真值:由仿真模型及给出的n(k),生成n(k+1)观察两种方式得到曲线的走向3、给出不同的值,观察n(k+1)随时间变化的情况判断标准如果要保持系统的稳定,至少n(k+1)应该保持在一个恒定的状态,或者逐渐趋于零。如果n(k+1)不断增加,则系统最终将趋于饱和,无法再接纳新的数据包,此时系统不稳定。仿真结果:1、验证在n的佔计误差较大的情况下的收敛特性:1)n=170;估计nt=20;m=100:20.2:N=100016030040050060070080g001CC08002)n=50;估计nt=180;m=1000=-1-02:N=80结果说明可以看出,当估计值与系统本身的积压数据包数有很大差别的时候,无论是大还是小,最终都可以趋于实际值,从而收敛特性得到验证。1)同时可以看到,改变的值:当λ增大的时候,收敛地更快;2)当n不变的时候,改变m的值,如果n/m变大,那么发生碰撞的几率就变大,也会导致估计的n值更快地趋向理论n值这些都是于课堂分析的理论情况相吻合的6/82、验证系统的稳定性下图分别为A=02:=10.1:=1:4=1+02:=0.3时候的情况。可以看到,当λ

无约束最优化：主要是共轭梯度法和拟牛顿法中DFP方法，很简单的程序。清晰易懂。初学者使用。