Warning: The solutions are parameterized by the symbols: z

Hi There

This is my first time posting, as i am new to matlab, so apologies if i have missed some etiquette that i know there is on some forums.

i am trying to calculate some values for angles of a mechanical arm with a linkage system. i have drawn out the system and found all the relationships i possibly can using the cosine rule and put them all into the file, hence some are not applicable to this bit of code.

my problem is i am trying to use the solve function to solve the following equation for zeta;

k2=sqrt(s^2+h^2-2*s*h*cos(theta+zeta))

i can find the solution of this using wolframalpha as;

zeta=-acos((h^2-k^2+s^2)/(2*h*s))-t

however if i write solve(k2, zeta) it comes up with the following error;

 Warning: The solutions are parameterized by the symbols: z7.
 To include parameters and conditions in the solution, specify the 'ReturnConditions' option. 
> In solve>warnIfParams (line 500)
  In solve (line 356) 
Warning: The solutions are valid under the following conditions: (in((theta + z7 + acos((H^2 + 270400)/(1040*H)))/(2*pi), 'integer') | in((theta + z7 - acos((H^2 + 270400)/(1040*H)))/(2*pi), 'integer')) & -1 <= (H^2 +
270400)/(1040*H) & (H^2 + 270400)/(1040*H) <= 1 & 0 < z7.
 To include parameters and conditions in the solution, specify the 'ReturnConditions' option. 
> In solve>warnIfParams (line 507)
  In solve (line 356)

I know that the answer for theta and zeta are 0.271 and 0.757 at the point where l is 330, i could just write in what the rearrangement is but that kind of defeats the object of having a maths package in my eyes and i would like to know for future knowledge what i'm doing wrong.

Hopefully someone can help me by telling me i'm an idiot and i've missed something stupid.

Ross

[Ross Hanna]

ok, so i've got past the first problem which is getting matlab to give me an answer other than "z". i did this by making theta and zeta "real" when writing syms. the problem now is its giving me two answers, one positive, one negative. i have tried B=A(A=>0) but it doesnt work. i have tried "assume" but it comes up with the error message

Undefined function 'assume' for input arguments of type 'logical'

any help would be greatly appreciated. thanks

Answer 1

[Andrew Newell]

The function solve is part of the Symbolic Math Toolbox. Are you defining the symbolic variables? If I run this code

syms s h theta zeta
k2=sqrt(s^2+h^2-2*s*h*cos(theta+zeta));
zeta = solve(k2, zeta)

I get

zeta =

 - theta - acos((h^2 + s^2)/(2*h*s))
   acos((h^2 + s^2)/(2*h*s)) - theta

[Ross Hanna]

Thanks for replying. Yes i am defining symbolic variables, not sure if this is the best way but i don't know their values without consulting a 3d model which i am trying to recreate and this is the only way of calculating them.

my problem is that when i run the program it gives me the answers above, but then lower down it has the issue of the error below;

Error using mupadengine/feval (line 163) The number of equations exceeds the number of indeterminates. This is allowed only for polynomial systems.

this is i'm guessing as sol_zeta is giving two outputs. how do i limit solve to one output? thanks

[Andrew Newell]

If this problem is well formulated, you could try both values of zeta, one at a time, and then determine whether one of them can be ruled out by the mechanical constraints. If not, you may need to think about the physics some more!