vpa(d,3)
We can tell from the number of decimal places that you show that you must be using symbolic numbers, so you can use vpa() to evaluate to a specific number of decimal places.
There is no way to round double precision numbers to a particular number of decimal places, because 0.1 is not exactly representable as a finite binary floating point number, for the same reasons that 1/7 is not exactly representable as a finite decimal number.
