Disclosed is a process lubrication optimization method aiming at stable rolling under large deformation of a DCR unit. The method mainly includes the following steps executed by a computer: (1) collecting main equipment and process parameters of a secondary cold rolling mill unit; (2) defining the parameters and intermediate process parameters; (3) initializing k' to 0; (4) calculating emulsionconcentration c; (5) initializing k to 0; (6) calculating an emulsion flow rate w; (7) initializing k to 0; (8) calculating an initial temperature t<0>; (9) calculating a friction coefficient mu; (10) calculating rolling pressure P and a forward slip value S; (11) calculating a function G<3>(Z); (12) determining whether inequalities are true; and (13) outputting optimum emulsion concentration c, an optimum flow rate w and an optimum temperature t<0y>. On the premise of stable rolling, the secondary cold rolling mill unit outputs the optimum emulsion concentration c, the optimumflow rate w and the optimum temperature t<0y>. Therefore, the cooling capability of an emulsion is fully developed, and the cooling capacity of the emulsion in each rack is more uniform, which improves the stability of production.