基于次优解群随机行为的复杂空调冷冻水管网能源利用效率优化研究

The key technique which researches initiative energy consumption is a model which accurately calculates thermodynamics characteristics of complex chilled water system. Thus, the characteristics of multi-variable, strong-coupled, nonlinear for chilled water system make its optimal coordinated control complexly diverse under partial load. The project proposes a suboptimal control method to study the thermodynamics suboptimal model building method about the air conditional system cold and humidity load thermodynamics suboptimal calculation in gridiron nodal points under thermodynamically stable chilled water system, and the relevant between the suboptimal solution of energy utilization efficiency of pipe network and variable of optimal control under selectively partial intervention of space distribution of humidity and cold load. And the method explored the randomness of suboptimal solutions in the different load distribution of complex chilled water system and revealed optimum control mechanism for energy utilization efficiency of pipe network which initiatively developed energy saving potential and also revealed the adaptation under different topological structure and load distribution of pipe network. There is a great significance for researching optimum control mechanism with selectively part intervening in complex chilled water system. It provides a new way to explore the optimization calculation and analysis method of thermodynamic model under the control of multi-variable, strong-coupled, nonlinear air-condition system.

复杂冷冻水系统热力学精确计算模型是研究其主动节能技术的关键，多变量强耦合的高度非线性特点决定了部分负荷条件下的冷冻水管网优化控制的复杂多样性，本项目提出次优控制方法，研究大型冷冻水系统在热力学稳定约束条件下，空调系统冷/湿负荷在管网节点空间分布控制的热力学次优化计算模型构建，以及冷/湿负荷空间分布选择性局部干预下的管网能源利用效率次优解与优化控制变量的关联性，分析不同负荷分布的复杂冷冻水系统次优解群随机行为，揭示管网能源利用效率次优控制机理，及其在不同管网拓扑结构与负荷结构条件下的适应性，主动创造条件以挖掘节能潜力。研究复杂冷冻水系统选择性局部干预控制下的优化控制机理，为探索空调系统多变量强耦合条件下热力学模型优化计算、分析方法及主动节能技术提供新的思路。

城市公共建筑空调系统普遍存在运行管理水平低、能耗高、过度维护等特点，能源、资源和人力浪费严重，但影响空调系统的变量类型复杂，数量庞大，属于典型多维度优化问题，本项目构建复杂中央空调冷冻水管网不同运行工况下普遍适应的精确热力学计算模型，研究管网负荷结构对其运行稳定性的影响规律，以冷（湿）负荷空间分布及管网拓扑结构为控制变量构建了管网能源利用效率次优化控制模型，采用多维度向量空间随机走步搜索的次优化计算方法，研究管网能源利用效率优化机理和可行性，研究结果表明，利用随机走步搜索的近似优化方法，能够在有限可控迭代计算量约束下，大大提高优化计算效率，尤其对于管网各节点阻力特性优化问题呈现极好的复现性和可靠性，可实现多达数十甚至数百个控制变量同步优化，其次优解群呈现显著的统计学特性，次优解群统计学特性为研究管网热力特性敏感性提供手段，并为进一步研究多变量约束的管网能源热力学特性、优化设计、运行大数据及能效诊断等科学问题研究奠定了理论基础和技术手段。