Constraining a variable value

I am constraining constants for the dynamics of an autonomous system that I am simulating. Within a for loop of 100 iterations, I would like to give "a" a random value greater than zero, "b" a random value greater than 1/a, and "c" a random value greater than 0.5. I have an idea for the code of "a" but I am confused on how to constrain "b" and "c" so that they are random values specifically greater than 1/a and 0.5. Here is what I have for my code right now:

AS= rand(1, 200)*1000;
for ii = 1:100 
    a = AS(ii);
    b=1/a;
    c=???
end

Any suggestions on this issue are greatly appreciated. Thank you.

[Steven Lord]

You've told us the lower limits for the values the variables can take on, but what are the upper limits? Is it 1, Inf, a function of one of the other variables like b's lower limit is a function of a, etc.?

Also what distribution do you want to draw the values from?

[Matthew Tortorella]

There is no real upper limit, but I just found it apprpriate for context of simulation to not go above 200. I can adjust that as need be. I would prefer to get random values from a normal distribution.

[Image Analyst]

@Matthew Tortorella, are you ever going to try my suggestion below?

[Matthew Tortorella]

@Image Analyst I tried it and it seems right for "c" but I need "b" to be greater than 1/a not greater than 0.5. Also can you explain what the c = a + (b-a) * rand equation is doing to create a random value for "c".

[Matthew Tortorella]

@Image Analyst I also just took another look at my Lyapunon derivative equation which is where I drew my conclusions about the retsrictions on a,b,c and they have changed. They are now: a>0, b>1/a, and c>1/2b. So the same except for "c" now.

Answer 1

[Image Analyst]

b = 1/a;
b = max([b, 0.5]); % Make sure b is > 0.5
c = a + (b-a) * rand;
c = max([c, 0.5]); % Make sure c is > 0.5