随着能源互联网的不断发展，城市规划的复杂度提升，人们对综合能源系统的探索越来越深入，而热电耦合系统作为最为典型的综合能源系统，将人类两大需求能源联合，提供相互独立热力系统与电力系统协作联合运行方式，打破了两个系统间的物理和信息壁垒。作为与电力系统结合的主要能源形式，区域热在多能流系统中发挥着关键作用，它的供热灵活性以及可利用热源的种类多样性是我们研究的动力。然而现有的热网模型研究对于网络热损失的处理方式有些单一，以及对于用户的热需求的定义和考虑不充分，这些问题限制了热电耦合系统的多样决策与效率提升。本论文研究旨在解决热电耦合系统中，单热源及多热源复杂热网络的物理建模问题，以及对热电耦合方式的创新设计，其主要贡献为：（1）在树状拓扑结构的单热源区域热网中，提出了一个稳态热流优化问题，量化用户侧的热需求，旨在使建筑室内温度和用户设定点之间的总偏差最小，以及在实际操作的约束下，热源的总输入能量最小。由于该优化问题是非凸的，提出了一个准确的、基于物理的热网模型来凸化并且有效解决这个问题。仿真结果验证了所提出的方法的有效性，该方法可以处理有损网络的复杂性并给出最优热流解决方案。（2）在多热源的复杂区域热网建模中，基于单热源热网的物理模型，成功推广泛化到多热源区域热网，该模型被证明仍然是凸优化模型，旨在降低总输出热能的同时满足用户热需求。提出不同热源厂之间的合作供水方式，使得系统的运行逻辑更加完整和实用。（3）基于水泵耗电送水的创新耦合方式，提出了热电分层耦合网络模型，以及追踪电能耗的优化方式，将多热源区域热网系统的模型应用到该耦合系统中，证明优化问题是凸问题，并利用分布式算法进行求解达到实时控制。综上所述，本文研究面向热电耦合系统中树状拓扑单热源及多热源的区域热网的建模，以及基于水泵的分层耦合方式，通过模型和运行逻辑的创新，解决了已有模型对拓扑结构和耦合机制的过于简化。研究成果可应用于能源互联网中广泛存在的热电耦合综合能源系统，为实现高效的多决策的最优能流计算、实时控制提供可实行的方案。

This thesis aims to solve the physical modeling problems of complex district heat networks with single source and multi sources in the integrated system, as well as the in- novative design of the coupling methods. The main contributions are: (1) A steady-state heat flow optimization problem is proposed in a single heat source district heat network with a tree topology to quantify the heat demand on the user side, aiming at minimiz- ing the total deviation between the building indoor temperature and the user set point, as well as the total input energy of the heat source under the constraints of practical oper- ation. Because this problem is nonconvex, an accurate and physics-based model of the (lossy) heat network is proposed to convexify and efficiently solve it. (2) In the model- ing of complex district heat networks with multiple heat sources, the physical model of the single-source heat network is successfully extended and generalized to a multi-source district heat network, which is proved to remain a convexity. A cooperative water supply approach between different heat source plants is proposed to make the system operation logic more complete and practical. (3) Based on the innovative coupling approach of wa- ter pumps consuming power to deliver water, a hierarchical combined network model for heat and electricity is proposed, as well as an optimization approach to track electrical en- ergy consumed by pumps. The multi-source heat model is applied to the coupled system, the optimization problem is proved to be convex, and the distributed algorithm is used to solve it to achieve real-time control.In summary, this paper studies the modeling of district heat networks with tree topol- ogy in combined heat and electric power systems, and the hierarchical coupling approach based on water pumps. Through the innovation of model and operation logic, the existing models’ oversimplification of topology and coupling mechanism is solved. The research results can be applied to the widely investigated integrated energy systems in the energy internet, and provide implementable solutions to realize efficient multi-decisional optimal energy flow calculation and real-time control.