随着南方区域电力市场建设的推进，跨省跨区电力交易规模逐渐扩大，未来计划和市场双轨制并存的格局将长期存在，其中合约电量的分解对市场运行和市场主体利益有重要意义。为了更好地促进发电商调度计划的实施、保证市场主体利益，提出了一种基于来水多区间不确定集的梯级水电站优先发电计划电量两阶段分解方法。在第一阶段的电量分解中，首先考虑丰水期和枯水期的来水特性，建立了丰水期-枯水期年度电量分解模型，实现年度电量到月度的“应季性”分解。然后以发电进度一致性为目标建立了基于进度系数一致性的月度电量分解到日模型。在第二阶段的电量分解中，考虑来水不确定性并将其描述为多区间不确定集的形式，从月度电量分解到日时间尺度入手，采用双层模型求解出第一阶段电量分解曲线覆盖下的保守电量分解曲线。最后，以云南省梯级水电站为例，分解得到了各电站的计划电量分解曲线和保守电量分解曲线，并对比分析了所提方法的有效性。

With the construction of the southern regional power market, the scale of cross-provincial and cross-regional power transactions is gradually expanding. In the future, the co-existence of the dual-track system of plan and market is expected to exist for a long time, in which the decomposition of the contracted power is of great significance to the market operation and the interests of the market entities. To better promote the implementation of the scheduling plan of the power generations and ensure the interests of the market entities, a two-stage decomposition method of the preferential power generation plan electricity based on multi-band uncertainty sets of inflow water is proposed. In the first stage of power decomposition, the annual power decomposition model for the wet season and the dry season is established by taking into account the characteristics of inflow water during the wet season and the dry season, so as to realize the seasonal decomposition of the annual power to the month. And then, with the goal of consistency of the power generation progress, a monthly power decomposition to daily model based on the consistency of progress coefficients is established. In the second stage of power decomposition, the uncertainty of inflow water is considered and described as a multi-band uncertainty set, and a two-layer model is used to solve the conservative power decomposition curve under the power decomposition curve of the first stage, starting from the time scale of the monthly power decomposition to the day. Finally, the planned power decomposition curves and conservative power decomposition curves of each station are obtained by decomposing the hydropower stations in Yunnan Province as an example, and the effectiveness of the method proposed is comparatively analyzed.