This preview shows pages 1–3. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: ME201 Advanced Dynamics (Fall 2007) HW3 Solution ( 100 pts ) 1. 50pts Consider the following linear dynamical system on R 2 : x = ax + by y = cx + dy In matrix form, this system can be written as x = A x for a matrix A , where x = ( x,y ). Find the matrix A . Find the eigenvalues and generalized eigenvectors (if they exist) for any value of ( a,b,c,d ). Draw the corresponding phase portraits. Show that the nature of the phase portrait is determined by Det ( A ) and Tr ( A ), the determinant and trace of the matrix A . In Det ( A ) Tr ( A ) plane, draw the various phase portraits. Do not forget the lines where you have zero eigenvalues. If needed, use help of numerical computation of trajectories. The matrix for this system is: A = a b c d The eigenvalues ( i ) and eigenvectors ( v ) are: i = 1 2 a + d p 4 bc + ( a d ) 2 v = a d 4 bc +( a d ) 2 2 c 1 ! where it should be clear that there are two eigenvalues and two eigenvectors. These results are those of Mathematica, they are specific to the case of nonzero entries (ie c 6 = 0). There are probably many different forms and normalized versions if you consider all cases. Along these lines consider the special case of A = a d where a = d . The two eigenvalues are clearly equal to a in this case and therefore we need generalized eigenvectors. Using the technique ( A 1 I ) x = we have the conditions: x 1 + 0 x 2 = 0 x 1 + 0 x 2 = 0 Because of this ambiguity we define two parameters = x 1 and = x 2 . Using this and the equations above we have x 1 x 2 = 1 + 1 for any nonzero and . So therefore the generalized eigenvectors for this particular case are 1 and 1 . To characterize the system with respect to its Trace and Determinant lets see how they are related. By definition for this two by two matrix: Tr ( A ) = 1 + 2 Det ( A ) = 1 2 ME201 Advanced Dynamics 1 HW3 Solution Since we have a good concept of phase space characteristics with respect to eigenvalues, lets solve for them using the two equations above: 1 , 2 = 1 2 Tr ( A ) p Tr ( A ) 2 4 Det ( A ) From here we note that qualitative changes in the properties of the eigenvalues occur when Tr ( A ) = 0, Det ( A ) = 0, Tr ( A ) 2 = 4 Det ( A ). We therefore investigate the regions of the Tr ( A ) vs. Det ( A ) plane which is subdivided by these conditions. The following two figures illustrate both this subdividedwhich is subdivided by these conditions....
View Full
Document
 Fall '08
 Mezic,I

Click to edit the document details