在“双碳”目标和能源转型的推动下分布式光伏以“点多面广”特性接入配网，使得配网层级间功率交互更加复杂，传统的供给侧调节方式难以满足多源协同与多目标协调的运行需求。为此提出一种考虑动态舒适区间的配网-台区-用户多层级多目标协同优化方法。首先，构建了配网-台区-用户多层级结构及分层调度框架：将配电网划分为配网层、台区层和用户层，各层源网荷储资源在本层内实现协同互动，层与层之间进行双向功率传输和信息交互，促进资源的合理分配与利用。其次，建立了考虑动态舒界动适区间的柔性资源模型，量化了居民用户在不同电价下的资源调控潜力，实现了调控边态更新。然后，建立了配网-台区-用户多层级多目标协同优化模型，综合考虑不同层级优化目标，制定自下而上的协调优化机制。最后，通过算例仿真对所提策略进行了验证，结果表明，所提方法能充分挖掘和利用各层级源网荷储资源，有效提高了光伏消纳、优化负荷曲线、降低用户用电成本以及保障配网安全经济运行。

Driven by the "dual carbon" goals and energy transition, distributed photovoltaic systems with their characteristics of "numerous points and extensive coverage" are connected to distribution networks, making power interactions between distribution network levels more complex. Traditional supply-side regulation methods struggle to meet the operational demands of multi-source collaboration and multi-objective coordination. To address this, a multi-level multi-objective collaborative optimization method for distribution networks-substation areas-users is proposed, considering dynamic comfort zones. Firstly, a multi-level structure and hierarchical scheduling framework for distribution networks, transformer areas, and users is established. The distribution network is divided into the distribution network layer, transformer area layer, and user layer. Within each layer, source-grid-load-storage resources achieve collaborative interaction, while bidirectional power transfer and information exchange occur between layers, promoting rational resource allocation and utilization. Secondly, a flexible resource model incorporating dynamic comfort zones is developed to quantify residential users′ resource regulation potential under varying electricity prices, enabling real-time adjustment of regulation boundaries. Then, a multi-level multi-objective collaborative optimization model for distribution networks-substation areas-users is formulated, comprehensively considering optimization objectives across different levels to establish a bottom-up coordination mechanism. Finally, the proposed strategy is validated through case simulations. The results demonstrate that the method effectively maximizes and utilizes source-grid-load-storage resources across all levels, significantly improving photovoltaic integration, optimizing load curves, reducing user electricity costs, and ensuring safe and economical distribution network operation