作为消纳新能源的有效方式之一，微电网是新型电力系统的重要组成部分。本文的研究以高比例新能源注入的微电网为背景，针对新能源功率预测和微电网优化调度目标不一致的问题，建立了以微电网运行效益最高为目标的功率组合预测与微电网日前、日内调度端到端优化模型，并提出了求解优化问题的方法。首先构建了双层优化问题，上层为功率预测模型训练，下层为微电网运行成本最小，将上层设计为组合预测问题，下层的运行成本优化结果设为组合权重优化问题的损失函数；然后通过启发式算法迭代求解上下层问题，解得使运行成本最低的预测结果和调度方案；最后在由IEEE33节点和IEEE123节点系统扩展的微电网中接入真实的新能源数据，验证了提出方法对于提升微电网运行效益的有效性。

As one of the effective methods for integrating new energy sources, microgrids are an essential component of the new-type power system. This paper aims to minimize the operational costs of microgrids under high injection of renewable energy sources, establishing issues related to the forecast combination of renewable energy and microgrid operation optimization, and proposing methods to solve these issues. Initially, daily-ahead and intra-day dispatch models for microgrids considering prediction errors were developed. Subsequently, the problem of finding the forecast results that minimize costs was transformed into solving for the combination weights of various forecasting sub-models, and heuristic algorithms were employed to solve for the combined forecasting weights. Finally, by integrating real renewable energy data into a typical microgrid extended from the IEEE 33-node system, the proposed method was validated for its effectiveness in enhancing microgrid operational benefits, addressing the challenge of discrepancies between renewable power forecasting and microgrid optimization objectives.