Modeling Vibrating Beam - North Carolina State University模拟振动梁-北卡罗来纳州立大学.pptVIP

Modeling Vibrating Beam -using the harmonic Oscillator equation verses collected data Initial Data Collected (unmodified) First Attempt to Model Data First Attempt Cont. Used fminsearch with initial conditions on C0=1 and K0=2 resulted in C=-.1163 and K=-4.4527 How can we improve on this model? Take out the zeroing out area. Shift graph to time=0. Rerun fminsearch. Taking Out Zeroing Area Rerunning fminsearch CO=1 K0=2 cost=6.3941x10^-7 C= -.67926355551960 K= -1.5227892111787423x10^3 But are these initial conditions the best ones to estimate C and K? Finding better estimates of C and K Rewrite code with double ‘for’ loops to check initial conditions between -5 and 5 for both C0 and K0 Stored CO and KO conditions with cost values into a matrix. This program ran overnight and did not finish in time. Checked the resulting (and incomplete) data for the smallest cost. The Better C and K We were able to look at 633 additional different initial C0 and K0 conditions. Found C0 = -5 and K0=-2.41 to have a lower cost of 6.3875x10^-7. This cost is lower than for C0=1 and K0=2 so we used the resulting C and K of… C = -0.68407341326426 K =-1.522633494063630x10^3 Determine Standard Error and Confidence Intervals for C and K Model fitted: d^2 y / dt^2 = - C dy/dt - K y n = 5019 C = -0s.e.= 095 perc. ci : ( -0 -0) K = -1522.633494 s.e.(K hat) = 0.383006 95 perc. ci : ( -1523.399506, -1521.867482 ) sigma^2 = 1.2666e-010 What do our estimates tell us? A closer look Looking at the differences The residuals gives us the difference between our data and the values estimated by the model. From the stats.m program that we were given, we have a graph of the residuals vs. time, and residuals vs. fitted data. Residuals vs. Fitted Value Residual vs. Time Sample Data vs. Standard Normal Deviation From Normal Distribution So what could we do better? A final technique to better model our data was introduced to our group by Profes