A multidiscipline collaborative optimization approach with acceleration strategies for generator maintenance scheduling in large hydrothermal power systems considering uncertainty of natural inflows

Abstract

Generator maintenance scheduling (GMS) is an important task in the dispatch and operation of hydrothermal power systems. On a long-term timescale, the uncertainty of natural inflows makes GMS essentially a stochastic optimization problem. The sample average approximation (SAA) method is used to convert the stochastic hydro and thermal generator maintenance model into a large-scale deterministic optimization problem, but it is difficult to solve. This paper proposes a multidisciplinary collaborative optimization approach with acceleration strategies (MCO-AS) for GMS, where the deterministic optimization problem is decomposed into a system-level optimization problem and discipline-level optimization subproblems through decoupling of the nonanticipative constraints and the maintenance variable consistency constraints. Three acceleration strategies, replacement and linearization, information pooling, and parallel calculation, are proposed and incorporated. Finally, computational experiments on a modified IEEE 30-bus system and a real large hydrothermal power system are conducted, and the results illustrate the effectiveness of the approach and acceleration strategies proposed in this paper compared to the traditional SAA method as a benchmark.