visualization of 3-dimensional system of differential equations
TRANSCRIPT
Yuanxun Bill Bao Simon Fraser University
July 9th , 2009
7/9/2009 Yuanxun Bill Bao Simon Fraser University
} System of Ordinary Differential Equations } Direction Fields } A 2-D example } Motivation } Arrows of the direction field } 3-D problem } Animate curves } Animate fields } More examples
7/9/2009 Overview 2
} System of 1st order ODEs } Autonomous system } Applications ◦ Dynamical systems ◦ Physical and Biological
models
7/9/2009 System of ODEs 3
} A graphical representation of the solutions of the 1st order ODE system
} Qualitatively visualize the solutions
} Very helpful when an analytical solution is hard to achieve
7/9/2009 DirecDon Fields 4
7/9/2009 A 2-‐D example 5
7/9/2009 A 2-‐D example 6
} 3-Dimensional system of 1st order non-linear ODEs
} In general, hard to solve
analytically
} Use direction fields to get a qualitative visualization of the solution
7/9/2009 MoDvaDon 7
7/9/2009 MoDvaDon 8
} Let’s try DEplot3d command in Maple
} For a better visual, needs to be 3D } Computationally efficient } Maple PLOT3D data structure ◦ PLOT3D(objects, options) ◦ square:= hfarray([[0,0,0],[1,1,0],[1,1,1],[0,0,1]]): ◦ PLOT3D(POLYGONS(square), COLOR(RGB,1,0,0),
AXESSTYLE(BOX)); ◦ Picture
7/9/2009 Arrows of the DirecDon Fields 9
7/9/2009 Arrows of the DirecDon Fields 10
7/9/2009 Arrows of the DirecDon Fields 11
} Make a 3D grid over the domain
} Calculate the direction at
each grid point
} Compute each arrow centered at its corresponding grid point
7/9/2009 Back to 3D case 12
} Make a 3D grid over the domain
} Calculate the direction at
each gird point
} Compute each arrow centered at its corresponding grid point
7/9/2009 Back to 3D case 13
} To get a better visual, consider placing the arrows RANDOMLY over the domain
} Still missing something? ◦ Speed of direction fields
} How to resolve this? ◦ Scale the magnitude of the
arrows ◦ Apply a color scale on the
arrows
7/9/2009 Back to 3D case 14
} To get a better visual, consider placing the arrows RANDOMLY over the domain
} Still missing something? ◦ Speed of direction fields
} How to resolve this? ◦ Scale the magnitude of the
arrows ◦ Apply a color scale on the
arrows
7/9/2009 Back to 3D case 15
} Static visualization is not enough to fully exhibit dynamical behaviors
} By using animation, the dynamics of the I.V.P
are recovered
} Animate the solution curves
} 3D example
7/9/2009 Animate Curves 16
} Model the direction fields as a “School of Fish”
} What trajectory should each “fish” follow ?
} animation
7/9/2009 Animate Fields 17
} Frame-by-frame animation } In the 1st frame, initialize the
direction fields } Treat each point where the
arrow lies on as the starting point of I.V.P
} Solve the I.V.P for each point
numerically } Compute the arrows for each
frame
7/9/2009 Animate Fields 18
} To have a smooth animation, at least 25 frames are needed
} If we have a direction field of 100 arrows, run
the animation requires computing at least ◦ 25 × 100 = 2,500 arrows OR ◦ 2500 × 18 = 45,000 points (square-base arrow) ◦ 2500 × 48 = 120,000 points (hexagon–base arrow)
7/9/2009 Animate Fields 19
7/9/2009 Rössler System 20
} animation
} animation
7/9/2009 Lorenz System 21
} animation
7/9/2009 Chemical Oscillators 22
} animation
7/9/2009 Future Enhancements 23
1. M.B. Monagan, K.O. Geddes, K.M. Heal, G. Labahn, S.M. Vorkoetter, J. McCarron and P. DeMarco, Maple 10 Advanced Programming Guide. Maplesoft, Waterloo, Ontario, Canada, 2005.
2. Richard H. Enns, George C. McGuire, Nonlinear Physics with Maple for Scientists and Engineers. Birkhäuser, Boston 1997
3. C. Henry Edwards, David E. Penney, Differential Equations and Boundary Value Problems, Computing and Modeling, 2008
7/9/2009 References 24
7/9/2009 Thanks Note 25