For the pricing in the non-convex electricity markets, the various pricing properties including supporting a competitive equilibrium, non-discrimination, cost recovery of participants, and revenue adequacy of the market operator can hardly be simultaneously fulfilled. Current pricing methods cannot balance these properties because there lacks a computationally tractable pricing model that explicitly formulates these properties as objective functions or constraints. In this paper, we propose a pricing optimization model taking prices as decision variables and taking important pricing properties as objective functions or constraints. Based on duality theory, the Karush-Kuhn-Tucker conditions of the self-dispatch problem processed by convex relaxation are extended to construct the constraints related to supporting a competitive equilibrium. Taking advantage of the modeling flexibility, the constraints for non-discrimination, cost recovery, and revenue adequacy are established. The delicately designed pricing optimization model is a linear programming problem with an acceptable computational burden. It applies to the market clearing problem with non-convexity, time-coupling and network transmission characteristics. To be specific, the application of the proposed pricing method for the unit commitment problem is provided. The reasonability, scalability, and tractability of the proposed approach are illustrated in systems with over 2000 buses. Numerical experiments show that the proposed method shows better performance compared with other representative pricing methods.