Research On Power System Unit Commitment With Dynamic Security Constraints

Security and economization are the fundamental requirements of modern power system. Under premise of security and stability, optimal operation has huge economic benefits. The unit commitment is an important problem in the field of optimal operation. Its task is to find an optimal schedule and a production level for each thermal unit over a time horizon in the power system daily operation plan. Mathematically, unit commitment is considered as a high-dimension, nonconvex, mixed-integer nonlinear programming problem. Commonly, it is very difficult to find the optimal solution in an acceptable time.On the basis of the existing achievements in the field of unit commitment, this dissertation proposes a novel parallel algorithm to solve for dynamic security constrained unit commitment. The involved dynamic security constraints include several important nonlinear constraints, such as power flow equations and transient stability constraints. The proposed algorithm employs an augmented Lagrangian method that involves the variable duplication technique and the auxiliary problem principle, so that the primal problem can be converted to its dual problem and get the separate structure of the augmented Lagrangian which can be easily implemented in parallel computers. Due to its robustness and computational efficiency, the predictor-corrector interior point method (PC-IPM) is introduced to solve the transient stability constrained optimal power flow (TSOPF) subproblem. The case studies on several test systems illustrate the proposed parallel computing algorithm is reliable, robust, computationally efficient, and promising to solve large scale unit commitment problems.