多重网络综合供能系统的随机优化调度及风险管理

With efficient utilization and saving of energy, the development of Combined Cooling, Heating and Power system (CCHP) can be promoted. It is an inevitable progress that Distributed Energy Sources (RES) (i.e. wind or solar energy, natural gas, biogas,…) should be connected to the system. The Multiple network, which is composited by Power supply network, the Pipeline Network Commissioning of the Cooling and Heating System, operational communications network and energy supplying units and loads, will take a fundamental change in operation and management of power systems and energy supply system. Due to the operation performance in Grid Connected and Islanding modes, the different operation manners of generating electricity, heating, cooling and combined energy from energy supply units, different controllable degree of load response, the phenomenon of extremely complex operating will occur. The action characteristics are analyzed for the distributed energy supplying unit, power, cooling, heating load, energy supplying network and the influence from information network. The equivalent conversion theory of the new grade utilization of multi-energy is explored. The hierarchical decomposition and coordination model are established for centralized dispatching and individual decision-making of energy supply system. The cummication delay and data lose are retrieved through the longitudinal transfer and feedback model. The network characteristics are taken into account under the differrent operation mode between Grid Connected and Islanding. The operation risks of energy generations are controled and constrained. Many calculation methods will be researched and developed, such as the Multi-stages Maximum Entropy Method for the upper and lower optimization problem, the run-of stream intelligent optimization algorithms for discrete variables, the stochastic optimization about improved α-superquantile with characterizing the uncertainty constraint. Innovative theories and basic method researchs for development of the smart energy networks are explored.

能源高效利用和节能减排推动电冷热综合供能系统发展，风光气等分布式能源并入系统是其发展必然，供电网、供冷热管网、调度通信网、供、用能单元所组成的多重网络使得电力系统和供能系统的运行管理发生根本性改变。项目针对供能系统联网与孤岛的不同运行模式、供能机组发电、供热、供冷以及联供等不同供能运行方式、负荷响应的不同可控程度等所带来的极其复杂的运行特性，分析分布式供能单元、电冷热负荷、供能网络特性以及信息网络影响等作用机理；研究新型电冷热能量梯级利用的等值转化理论方法；建立供能系统的集中调度和个体决策的分层式分解协调模型；利用纵向递阶传递和反馈模型处理通信延迟与丢失补偿；考虑电网不同运行模式下的网络特性；控制和约束供能单元随机运行风险。研究解决上下层优化问题的阶梯式极大熵算法、离散变量的径流式智能优化算法、改进超分位数刻划不确定约束的随机优化等算法。力图为智能能源网络的发展探索创新理论和基础方法研究。

随着综合供能系统的发展，风光气等多种分布式能源转化为电能及冷热能并入中和供能系统，电冷热供能系统的运行管理发生根本性变化。项目研究了系统优化协调与风险管理的理论和方法，主要包括下面四个方面的内容： 综合供能系统分层分解协调调度；多联供机组在供电、供热、供冷、联供等不同生产状态下具有能量品味差异的电能与冷热能的等值转化；对多重网络综合供能系统进行风险管理；处理不连续可导等无法满足数学收敛性要求等问题的智能优化算法。提出了以总供能周期内系统各供能单元的联合供能成本最小为目标，求解最优供能方案；采用“以热定电”和“以电定热”两种配网调度策略模拟热力管网优先、配电网优先的管理方案，解决计及冷热电多联供机组的配网经济调度；对响应峰谷电价的冷热电联供系统进行了优化调度。新建了考虑线路阻塞的风险限制调度多步整合模型。提出了风电以制热方式和供电方式参与冷热电联供系统供能的供热模式及供电模式，并用超分位数方法刻画风电随机性、波动性对供热优化模型及供电优化模型的影响，对风电并入冷热电联供系统的进行优化运行控制；提出了联供系统的能量流函数，提升了冷热电联供系统的经济环保效益；采用了条件风险价值的方法刻画随机因素对多重网络综合供能系统的影响，并进一步推导了多随机变量的超分位数方法；提出了矩不确定分布鲁棒优化方法，解决含风电出力矩不确定的电力环保经济调度问题；提出了一种解决三相负荷不平衡引起线损过高的用户相别优化调整方法，针对风电出力波动性和间歇性对电力系统运行的影响，构建了基于条件风险方法的电力系统旋转备用模型，并进一步建立了电力系统双层随机优化调度模型，优化火电机组出力及风电计划出力。针对标准粒子群算法局部最优和收敛过的问题，提出空间粒子群优化算法，降低了计算结果的随机性。针对传统粒子群算法早熟的缺陷，提出模糊自修正粒子群算法，增强全局搜索能力。针对多重网络综合供能系统优化调度的智能算法，研究了源于大自然径流规律的智能优化算法—径流算法，可求解非线性、非凸、不连续的典型优化问题。本项目的研究成果为实现大规模分布式电源高效利用、智能能源网络发展具有理论和现实意义。