摘要

Transient stability constrained optimal power flow (TSCOPF) is an important tool for power system planning and operation. In the paper, a discontinuous non-convex TSCOPF model with many minima is proposed to fully consider the generation valve-point effects and prohibited operation zones (POZs). Though the model is hard, it is general and flexible not only to accommodate any complex system components and discrete control variables, but also to be solved with any state of the art heuristic solution methods. Furthermore, an improved Group Search Optimization (IGSO) algorithm based on backward searching strategy, Cauchy mutation and inheritance operator is tailored to solve this challenging discontinuous non-convex problem. Four comprehensive case studies on the WSCC 9-bus system, New England 39-bus system, and IEEE 145-bus system have confirmed the validity of proposed discontinuous non-convex TSCOPF model and the outperformance of IGSO solution method compared with seven representative artificial intelligence algorithms.