Finding the mathematical model for a DIFFERENCED ARIMA

[TryingwithSAS]

Hi, I had to difference my dependent variable to get white noise autocorrelation. Below is my code. How do I get the mathematical form of my ARIMA model? I got several MLEs but when I multiply them by the variable values, I don't get the forecasted values. I need an equation: dependent var = mu + coefficien1*X1 + coefficient2* x2 + coefficient3*x3

No clue how to do this with a differenced response variable.

PLEASE HELP?
TXX

TITLE 'REGRESSION ANALYSIS 8'; run;
PROC ARIMA DATA = SAHT.XCOMBINED3 ;
IDENTIFY VAR = DEPENDENTVAR (1) CROSSCOR = (
X1 X2 X3  );
ESTIMATE P = (1 6) METHOD = ML INPUT = (
X1 X2 X3 ) PLOT;
FORECAST ID = TIME INTERVAL = WEEK PRINTALL OUT = SAHT.XOUT;
RUN;
QUIT;