【作者】 周守军；

【导师】 潘继红；

【作者基本信息】 山东大学 ， 热能工程， 2012， 博士

【摘要】 集中供热在国内的发展非常迅速,无论是供热规模、供热半径还是供热管网型式,都发生了巨大的变化。为了进一步提高集中供热系统的能源利用率,更好地发挥其节能减排的特点,本文提出了一种基于管网动态模型的城市集中供热系统运行参数预测及优化运行方法,并对其进行了系统的研究,验证了所提出的动态模型与算法的有效性与准确性。结果表明,该方法对推动集中供热系统运行调节与优化控制技术的发展,促进城市集中供热系统运行管理水平的提高具有指导意义。首先,基于网络图论及基尔霍夫定律,建立了集中供热管网水力模型,采用改进平方根法进行求解。针对水泵扬程取定值与采用拟合公式两种管网定压方式,对模拟一次管网进行了水力工况建模分析,结果表明初调节是保证管网水力平衡的重要前提。其次,根据能量守恒方程,推导出热水供热管道热力工况动态模型。针对该双曲型方程,采用特征线法建立其特征线方程。然后采用逆步进法、一阶泰勒级数展开及向前显式差分格式,沿特征线建立该方程对应的有限差分方程；提出采用最小二乘法拟合出不同管径、不同保温材料的供热管道散热损失与供水温度的线性方程,以便热力工况动态模型调用；基于以上内容,分别推导出单独考虑管道散热效应的热力模型,称为“散热模型”,以及同时考虑散热与蓄热的热力模型,称为“蓄热模型”。然后,分别建立了模拟一次管网的“散热模型”与“蓄热模型”。采用散热模型对某工况进行模拟,分析了不同误差限、不同空间步长以及不同初值对模拟结果的影响。为了保证差分格式的收敛,采用各计算节点两次迭代温度差与当前时间步长的比值,即温度梯度代替温度差,作为判断程序是否收敛的条件,效果良好；分别采用散热模型与蓄热模型对2种不同运行工况进行模拟,发现由于管网散热,导致远端用户入口温度降低,进一步加剧了远端用户与近端用户的热力失调,增加了管网运行调节的难度。蓄热效应使得管网在任何扰动下,趋于稳定的延迟时间增大,从而对确定管网进行调节的最佳时间产生重要影响。因此,建议在需要着重考虑管网稳定延迟时间时,采用蓄热模型；对于一般情况,应首先选用模型相对简单、收敛快的散热模型。针对集中供热管网节点泄露故障,采用模拟管网对3种不同泄露工况的流量与压力分布进行了仿真,即泄露点在供水管、泄露点在回水管以及供水管两点泄露,并对不同泄漏量进行了比较；针对第一种泄露工况进行了热力工况动态模拟,总结了管网泄漏时各点压力、流量及温度的变化规律,为管网泄露故障及泄露点位置的诊断提供相应的参考。建立了包含一、二次管网的间接连接联合管网动态模型。该模型包括1个一次模拟管网、10个二级换热站及二次网,每个二次管网包含10个,共计100个建筑用户(采用散热器模型整体描述)。采用该联合管网模型,对热计量下,用户自主调节阀门,控制室内温度情况进行模拟,模型良好地反映了用户处自主调节对室内温度、用户负荷以及一二次网运行工况的影响；选取某城市集中供热系统部分管网,采用散热模型建模。然后选择其中6个换热站半小时内的一次侧供水温度与供水压力监测数据平均值作为验证数据,与模型仿真数据进行比较。结果发现,除了1个小区换热站供水温度与供水压力相对误差超过3%以外,其余5个站的相对误差均在2%范围以内,验证了本文所提出的模型及算法的有效性与准确性。针对集中供热系统质量并调运行优化问题,提出建立以供热系统运行能耗费用最低为优化目标,以供水温度和供水流量为变量的集中供热系统一次网运行能耗费用方程。然后,根据非线性规划理论,应用一般约束乘子法将约束问题变为无约束问题,采用共轭梯度法,结合线性搜索,建立了该优化模型的求解算法及流程。最后,针对某城市集中供热系统一次网,将提出的非线性规划方法与逐步长搜索法相比较,结果证明该方法效率更高,结果更准确；应用本文提出的运行优化方法指导管网质量并调,与传统的质调节相比,节能效益明显,进一步验证了该方法的有效性。更多还原

【Abstract】 District heating rapidly develops in China. Whether district heating scale, radius or pipe network type, all have greatly changed. In order to further improve energy efficiency of district heating system and better bring its characteristic of energy saving and environmental protection into play, an operational parameter prediction and optimized operation method of district heating system, which is based on pipe network dynamic model, is proposed in this paper. This method is systematically researched, and the effectiveness and accuracy of the proposed dynamic model and algorithm have been verified. The results show that the proposed method has guiding significance to promote development of operational adjustment and optimal control technology, and to improve operational management level in district heating system.Firstly, based on network graph theory and Kirchhoff s law, the hydraulic model of district heating network is established, and solved by improved square root method. Aiming to two different constant pressure forms, pump lift adopts fitting formula and constant value, the hydraulic model of a primary simulated network is established. The results show that initial adjustment is an important prerequisite to guarantee the pipe network hydraulic balance.Secondly, according to the energy conservation equation, the thermal dynamic model of hot water heating pipe is derived. This dynamic model is hyperbolic equation, so that the characteristics method can be used to establish its characteristic equation. Then its corresponding finite difference equation along the characteristic line is set up by the inverse step method, first-order Taylor series expansion and forward explicit difference scheme. Then the least squares method is proposed to fit linear equations of pipe heat loss and water supply temperature for different diameter and different insulation materials, then these fitting equations can be used by the thermal dynamic model. Through the above, the thermal model that considers heat dissipation individually is deduced and called "dissipation model". At the same time, the thermal model that both consider heat dissipation and heat accumulation is deduced and called "regenerative model"Then, the dissipation model and regenerative model of the primary simulated network are established respectively. Through simulating a working condition by the dissipation model, the influence on simulating result is analyzed from error limit, space step and initial value. In order to ensure the convergence of the differential format, the temperature difference, which is the difference of two neighbor iterative temperature values on each computing node, is substituted by the temperature gradient, which is the ratio of the above temperature difference with the current time step, to determine whether the program is convergence or not, and it is proved that the method is good. At the same time, two different operating conditions are respectively simulated by the dissipation model and the regenerative model. It is found that the pipeline heat dissipation leads to lower the entrance temperature of remote users, and to further exacerbate the thermal imbalance and adjustment difficulty of pipe network. The regenerative effect makes the delay time of network from any disturbance to stabilization increase, then it will affect when the operational strategy is adopted to adjust the pipe network. Therefore, the regenerative model is suggested to be used when the delay time of network stability is more important. Otherwise, the dissipation model will be firstly selected because of its relatively simple and fast convergence.Aiming to the node leak failure of district heating pipe network, The flow and pressure distribution of three different leak conditions are simulated by the simulated network, which is one leak point in the supply pipes, two leak points in the supply pipes and one leak point in the return pipes, different leakage have been compared either. Furthermore, the thermal dynamic of the first leak condition are simulated. Finally, the pressure, flow and temperature variation of each point on the pipe network is summarized when the node leak happens, and it could provide corresponding reference for diagnosis of pipe network leak failure and leak point position.A dynamic model of an indirect connection pipe network is established, and it includes a simulated primary network,10secondary heat exchanger stations and secondary networks which consist of100building users. Each building user is overall described by a radiator model. This model is used to simulate the case that users independently regulate valve to control indoor temperature when the method of heat metering is adopt. The results show that the model well reflects the influence on the indoor temperature, the user load and the operating conditions of the primary and secondary network from the user’s independent regulation. Then, part of a real city heating pipe network is modeled by the dissipation model. And it is taken as validation data and compared with simulated data that the monitoring data average of primary supply water temperature and pressure in half an hour are collected from6selected heat exchanger stations. The results show that the relative errors of supply water temperature and pressure in5stations are all in the range of2%, in addition to one station whose relative error is more than3%. And the results verify validity and accuracy of the proposed model and algorithm in this paper.For the operational optimization problem of supply temperature and flow adjustment (known as "quality-quantity" regulation) in district heating system, an operational cost equation of primary network is set up, taking the supply water temperature and the water flux as variables. The optimization objective of this equation is to minimize operational cost. Then, based on the nonlinear programming theory, a constraint problem is translated into an unconstrained one by general constrained multiplier method, the conjugate gradient and linear search method are integrated together to solve the optimization problem. Finally, the presented method and the step-by-step method are respectively applied to a real urban district heating system. The result proves that the former is more effective and accurate than the later; the energy saving effect of quality-quantity regulation which is guided by the presented method is compared with the one of quality regulation, it proves that the former is prior to the later and further verifies effectiveness of the presented method.更多还原