In this paper, we propose a new consensus protocol in the form of distributed model predictive control (DMPC) in order to deal with the consensus problem of first-order discrete-time multi-agent systems. Equivalent form of general numerical solution of DMPC is presented by solving optimization problem, and necessary and sufficient convergence conditions of our proposed DMPC are obtained by using the algebra and matrix theories. New protocol with equivalent solution is beneficial to the theoretical analysis and avoids the highly computing cost. Numerical simulations demonstrate the effectiveness of theoretical results.