# Another second order problem

• May 19th 2011, 07:36 PM
ain
I still need help to convert another system to that form, but this one has four input (u1 ... u4). m1, m2 and m3 are scalar. The dynamic system is:

xddot = u1
yddot = (m1/m2)*tan(psi)*u1
zddot = (-m1/m3)*tan(theta)*sec(psi)*u1
phiddot = u2
psiddot = u4

Following Ackbeet's method in my other thread, and got like this pattern:
y1ddot = v1
y2ddot= v4
y3ddot = v3
y4ddot = v2
y5ddot = (m1/m2)y2v1
y6ddot = (-m1/m3)(1/sin(psi)) y2y3v1

May be the final solution can be simplify to be more close to the new form pattern, because in my calculation still has like sin(psi) in y6ddot. I want to check it with Ackbeet's calculation.

Thank you very much.
• May 19th 2011, 08:07 PM
Ackbeet
I need to know where you want to end up. I'd recommend letting some of your new LHS variables be equal to trig functions, like in the other thread.
• May 19th 2011, 08:25 PM
ain
I want to get the standard pattern like before (chained form)

y1ddot = v1
y2ddot = v2
y3ddot = y2v1
y4ddot = v3
y5ddot = y4v1
y6ddot = v4

Thank you.

Quote:

Originally Posted by Ackbeet
I need to know where you want to end up. I'd recommend letting some of your new LHS variables be equal to trig functions, like in the other thread.