A PSO-DMPC-based reaction regeneration system optimization control method disclosed by the present invention comprises the steps of S1 converting a transfer function model of a reaction regeneration system into a step response model; S2 establishing a DMPC model comprising an open loop prediction module, a steady-state objective calculation module and a dynamic matrix control module; S3 utilizing the strong randomness of the particles in the search space in a PSO algorithm to solve an economic optimization function within a larger range under the premise of not relaxing the constraint conditions; S4 according to the solution of the PSO algorithm to the economic optimization function, obtaining an output set value of the reaction regeneration system, taking the deviation of the output set value and the actual output as a target error function, and finally utilizing the PSO algorithm to solve the target error function, and obtaining an optimal variable quantity of an operation variable. The PSO-DMPC-based reaction regeneration system optimization control method provided by the present invention enables the hardware burdens of the reaction regeneration system (RRS) to be reduced, also can obtain the better operation variable parameters, and further optimizes and controls the RRS on the basis of guaranteeing the economic benefits.