Getting help with PIL simulation for an Arduino 32-Bit Board

Hello dear friends. These days I have got questioned about the PIL test using a 32-Bit Arduino Board. As you know, the 32-bit boards like mine, can be represented by single precision numbers. However, when we model an algorithm in a MATLAB Function in the Simulink environment, the default for the input, output and the intermediat variables is Double precision numbers. Now I wanted to know, In order to have a more precise PIL simulation, if I am supposed to convert the algorithm in the MATLAB Function to a Single Precision one?

An urgent and appealing request for help.

Thanks to all in advance.