In Simulink how would you change [4x1] signal to [1x1]

I would like to take a signal in simulink for example a =123 and b = 456 and have the output = 123456.

How would you do this?

Answer 1

[Walter Roberson]

Arithmetic blocks? One that multiplies by 1000 and the other that adds ?

Or is the range of values not set in advance? Is the range variable? For example if one time a = 123 b = 456 then you want 123456 as output; if the next time a = 123 and b = 45 then would you want 12345 as output, or would you want 123045 as output?

What result do you want if one of the values is negative? If both of them are negative then do you need a negative result?

[Adam s]

Walter,

if the next time a = 123 and b = 45 then would you want 12345 as the output.

[Adam s]

I will never have a negative number the lowest the values can go is zero.

[Walter Roberson]

https://www.mathworks.com/help/simulink/slref/mathfunction.html

https://www.mathworks.com/help/simulink/slref/roundingfunction.html

https://www.mathworks.com/help/simulink/slref/minmax.html

https://www.mathworks.com/help/simulink/slref/product.html

2   (constant block) ->
                      | max (minmax block) --> [T1]
                   B ->
                   
                   [T1]  ->
                          |  multiply (product block) -> log10 (math block) -> [T2]

(1+20*eps) (constant block) ->  


[T2] -> ceiling (rounding function) -> 10^u (math) ->
                                               | multiply (product block) -> [T3]
                                            A ->

 [T3] ->
       |    add (math)  --> result
   B  ->
   

In the above, I give an operation name, and then in () I give the name of the block that implements it.

In the above, the [T1] are not operations, but are simply labels to make clearer how the blocks join together

So you take

   A * 10^ceil(log10((1+20*eps)*max(B,2))) + B

The max() part is to handle 0 and 1 the same as 2 to 9.

The 1+20*eps part is a hack so that exact powers of 10 such as 1000 get the full number of digits, since ceil(log10(99)) is 2, same as ceil(log10(100)). The constant 20*eps is sufficient up to 1E32... but of course if you are above 1E15 then you have to start worrying about losing precision when you do the concatenation.

Answer 2

[Sameer Pujari]

You can do this

[Adam s]

Yes, techinally that works for the example given. However if the math is not the easy how would you do it?

Such as in matlab you can do:

a = 1;
b= 2;
c=3;
str = sprintf(%d%d%d, a,b,c)
double = str2double( str)
dobule = 123

How do you do this in Simulink?

[Sameer Pujari]

Then try to use Simulink function block

https://in.mathworks.com/help/simulink/ug/creating-simulink-functions.html