【作者】 施志钢；

【导师】 胡松涛；

【作者基本信息】 西安建筑科技大学 ， 供热、供燃气、通风与空调工程， 2009， 博士

【摘要】 我国的建筑能耗在社会总能耗的比例逐渐提高,在建筑的全寿命周期内,建筑运行能耗在建筑节能中处于重要的地位,重点是提高采暖、空调系统的能源利用率。土壤源热泵系统以节能、环保、可靠性高等优点,被称为是21世纪最具有发展前途的建筑节能技术之一,在世界范围内迅速发展和广泛应用。作为一项复杂的系统工程,土壤源热泵系统得到深入的研究,研究的领域主要集中在土壤换热器传热特性的模拟计算、提高传热性能、土壤源热泵系统运行特性的试验研究和土壤换热器的设计方法等,这些研究主要对系统的设计提供理论上的指导。尽管可以通过输入少量的电能获得较高的能量输出,然而在运行时,不论周围环境和设备本身特性如何变化,如果始终使土壤源热泵系统维持在最优状态运行,最大限度地挖掘系统的节能潜力,将能够大幅提高系统的性能,这就需要实现真正的优化控制。对于优化控制的关键是从本质上分析与能量特性有关的动态特性,基于此,本文针对土壤源热泵系统的优化控制进行了深入的理论和实验研究。土壤源热泵系统在运行过程中持续受到外界环境的扰动,而优化控制结构的设计和实现则强烈依赖于扰动的特性,对于土壤源热泵系统,按照扰动的频率可以分为两类：一类是与优化过程无关的快扰动；另一类是慢扰动,这类扰动将影响土壤源热泵系统运行的经济性指标。在充分分析土壤源热泵系统优化控制的可行性、特点及优化控制中所面临问题的基础上,基于大系统的分层递阶控制结构进行控制任务的分解与协调,提出适用于土壤源热泵系统的优化控制结构,该结构由调节层、优化层和自适应层组成,其中调节层抑制不可控的快扰动,并使控制变量维持在设定值；优化层根据经济性指标和相应的模型计算调节层的设定值；自适应层则适时地修正优化层的模型参数。这样的优化控制结构为进一步研究土壤源热泵系统的优化控制创建了理论框架。在基于模型的优化控制方法中,系统的过程模型是实现优化的基础,而系统过程模型由部件模型组成。土壤源热泵系统由地上的水-水热泵机组和土壤换热器组成,其动态模型作为自适应层的模型结构,对应的稳态模型为优化层提供必要的约束条件。对于水-水热泵机组,由蒸发器、冷凝器、节流机构和压缩机组成,蒸发器和冷凝器采用分段集中参数的方法建立动态参数模型,与压缩机和膨胀阀的稳态模型相结合,建立了完全由系统状态变量和输入变量表示的水-水热泵机组的数学模型。将仿真结果和试验数据进行比较,验证了模型的正确性。作为土壤源热泵系统的重要组成部分——土壤换热器,通过载能流体与水-水热泵机组相互耦合、相互影响。在深入分析土壤换热器模型的基础上,从有限长线热源模型出发,建立热渗耦合作用下的土壤换热器动态数学模型,采用格林函数法得到解析解表达式,通过解析解的分析,纯导热情况下的解析解只是一个特例。对于钻孔内部的传热过程,采用能量平衡法,建立钻孔内单U和双U支管内载能流体的温度分布模型,得到温度分布的解析解表达式,进而能够确定进出换热器温度。在水-水热泵机组和土壤换热器动态参数模型的基础上,提出利用实际存在的操作波动造成的系统动态变化过程辨识模型中的参数,根据辨识结果对优化层进行校正,使数学模型和实际系统由于失配造成的误差最小。根据最大似然估计将参数辨识问题转化为有约束的非线性规划问题,避免了复杂的数学推导,并可以把参数所具有的先验知识作为约束条件结合在这一框架之中。通过理论分析,证明了所建立的土壤源热泵系统模型存在唯一解及参数的可辨识性。对于循环水泵能耗模型,采用慢时变递推最小二乘法进行辨识。在全局搜索算法Scatter Search和局部搜索算法SQP的基础上,提出了一种新型的混合算法SS-SQP。通过分别采用Scatter Search算法、GAS算法和SS-SQP算法对水-水热泵机组模型参数进行辨识,结果表明SS-SQP算法充分发挥了Scatter Search算法和SQP算法各自的特点,使搜索性能获得了较大的提升。根据所提出的土壤源热泵系统的优化控制结构,将水-水热泵机组、循环水泵和土壤换热器模型耦合,从全局系统的角度建立土壤源热泵系统的运行优化模型,确立了目标函数、约束条件和控制变量。由于各部件之间相互耦合、相互作用,变量之间存在复杂的约束关系,属于有约束的多变量非线性问题。通过对优化模型的仿真计算,分析了钻孔壁温、回水温度和负荷率对最优运行工况的影响,并与定流量工况进行了比较。结果表明无论是在制冷模式还是制热模式,采用优化控制能够降低系统的总能耗,系统COP得到提高。对土壤源热泵统进行了实验研究,实测得到土壤的初始温度和土壤的综合导热系数,并验证土壤源热泵系统模型的正确性。在运行过程中,将各控制环路的设定值重置为优化值,对比优化前后的系统能耗及运行效果,结果表明采用优化控制后,系统运行稳定、可靠,系统的能耗降低,从而验证运行优化控制的可行性及可靠性,为土壤源热泵控制系统的设计和运行管理提供了重要的参考。更多还原

【Abstract】 The proportion of building energy consumption to total social energy consumption is increasing gradually. In the total life cycle of building, the operation energy consumption has an important role in building energy saving. Enhancing utility efficiency of HVAC is crucial to building energy saving. Ground source heat pump system (GSHPS), which has advantages of energy saving, environmental protection and high reliability, is regarded as an air conditioning technique having the greatest developmental future in the 21st century. So GSHPS is developed rapidly and applied widely all over the world. As a complex system engineering, it attracts more attention and many in-depth researches are implemented in the last decades. However, most of the studies focus on heat transfer of geothermal heat exchangers, enhancing heat transfer performance, designing method, simulation and testing of GSHPS and are mainly used as theoretical guidance in the designing of GSHPS. If continuously maintaining the GSHPS running at its optimum operating condition, despite the changes of environmental condition and equipment behavior, and digging the energy saving potential to the greatest extent, it is possible to achieve an important performance improvement. Continuous tracking and driving the process to its best operating conditions are termed as optimizing control. The essence of optimal control for GSHPS is to analyze dynamic characteristic with respect to energy. So Theoretical and experimental study on GSHPS optimal control is carried out in this dissertationGSHPS is continuously subjected to a variety of disturbances during operation. Design and implementation of optimizing control are strongly depended on the characteristics of the disturbances. The process disturbances of GSHPS are divided in two categories:one is fast varying disturbances which are irrelevant for the long term optimization of the process; another is relatively slow varying disturbances which have a significant impact on the optimum economic performance of the GSHPS. Thus the multilayer approach of hierarchical control theory based on the control task decomposion can be used to synthesis the control structure to reduce the mathematical complexity. Based on the analysis of feasibility and characteristics, a multilevel hierarchical control structure scheme for GHSPS including regulatory layer, optimization layer and dynamic adaptive layer is proposed. The regulatory layer takes care of the control tasks of keeping the controlled variables at given set-points, and suppress the influence of the "fast varying" uncontrolled disturbances. The optimization layer takes care of the optimizing control tasks, where the objective is to determine optimum set-points based on an appropriate performance criterion and a mathematical process model. The purpose of the adaptive layer is to compensate for model-induced errors by adjusting the model parameters. This control structure supplies the theoretic frame to further research.Process model of GSHPS is the foundation to realize the optimal control. GSHPS consist of water-to-water heat pump and geothermal exchanger (GHE), whose dynamic model act as the model structure in parameter identification phase. And corresponding steady model supplies necessary constraint condition for optimal layer. Water-to-water heat pump consists of evaporator, condenser, expansion valve and compressor in more detail. The spatially distributed heat exchanger including evaporator and condenser are partitioned into spatial zones in the model formulation stage, and lump parameter method model spatial zones based on the first principle. Coupled with compressor and expansive valve of steady model, an apace state model described by state variable and input variable is established. To verify the validity of modeling procedure of heat pump, the simulation outputs were compared with the corresponding values reported in reference papers and the results validate the accuracy of the model.GHE couple with water-to-water heat pump through energy carrier media and affect each other mutually. Under the assumption of the finite line source and considering the heat conduction and heat advection effect, mathematical GHE model is established. The analytical solution of temperature-field distribution is obtained by utilizing Green function method. The analytical solution under the condition of pure heat conduction is a special case. In the internal of bore well the thermal conduction of single U-tube and double U-tube is modeled and temperature distribution along the vertical of bore well is obtained using energy balance principle, thus the input and out temperature of GHE can be calculated.To achieve faster convergence to the optimum operating region, and to cope with persistent disturbances, a nonlinear non-steady-state process model is identified in the identification phase by change manipulative to make mismatch error between model and real system minimum. The identification result is transfer to optimal layer to revise optimal model to make optimal solution approach to the genuine optimal solution. According to the maximum likelihood estimate method, the parameter identification transform nonlinear program problem. Avoid complicated mathematical deduction. And a priori parameter information is included in the minimization criterion, thus this makes the estimation problem more robust with respect to missing process excitations and over-parameterization. Furthermore, by theory analysis the GSHPS model exit uniqueness solution and parameter identifiability are proved。Based on the improved scatter search algorithm and the local search SQP algorithm that fit continuity space optimization, a new mixing algorithm SS-SQP is proposed. Comparing to SS and GAS, the result of parameter identification shows that this new mixing algorithm can improve computing efficiency.According to the above optimal control structure, the operation optimal model for the global system is established, which the objective function, constraint condition and manipulate variable is defined. Each of the components is coupled and influence interactively. Optimization problem of GSHPS is nonlinear in both objective function and constraint conditions. The updated steady-state part of the process model is used to optimize the economic performance of the process, where new optimum set-points are calculated for the regulatory control system. Through simulating, effecting factors of bore well temperature, return water temperature and load ratio are investigated. Whenever operating under the heating or cooling mode, the total energy consumption can be decreased by the optimal control compared to constant water volume, and the system COP can be improved.Finally, experimental studies on GSHPS are carried out. Comprehensive thermal conductivity coefficient and initial temperature of soil are obtained. And then the GHSPS model is verified. To demonstrate the feasibility and reliable of optimal control, the set values in control loops are adjusted to their optimal values. Then the total system energy consumption is compared with that before optimization, and the result shows that the total energy consumption decreases after optimization, and the GSHPS can operate reliable and stably.更多还原