Unified solution of a non-convex SCUC problem using combination of modified Branch-and-Bound method with Quadratic Programming

In this paper, a new practical method is presented for solving the non-convex security constraint unit commitment (SCUC) problem in power systems. The accuracy of the proposed method is desirable while the shorter computation time makes it useful for SCUC solution of large-scale power systems, real-time market operation and long-term SCUC problems. The proposed framework allows inclusion of the valve point effects, warmth-dependent start-up costs, ramp rates, minimum up/down time constraints, multiple fuels costs, emission costs, prohibited operating zones and AC power flow limits in normal and contingency conditions.To solve the non-convex problem, combination of a modified Branch-and-Bound method with the Quadratic Programming is used as an optimization tool and a developed AC power flow algorithm is applied for considering the security and contingency concerns using the nonlinear/linear AC model. These modifications improve the convergence speed and solution precision of SCUC problem. In the proposed method, in contrast with traditional SCUC algorithms, unit commitment solution, checking and satisfying the security constraints are managed simultaneously. The obtained results are compared with other reported methods for investigating the effectiveness of the proposed method. Also, the proposed method is applied to an Iranian power system including 493 thermal units.

► A hybrid SCUC solution is developed to deal with large-scale, real-time and long-term problems.
► New formulations are proposed for considering valve point effect and warmth-dependent start-up cost.
► A new algorithm is developed for modeling the AC power flow in SCUC problems.
► Using the power flow algorithm both steps in traditional SCUC is done simultaneously.
► The proposed method provides better solutions than previous ones with a fast speed.