Abstract

The heating network of a big city is a complex hydrodynamic system consisting of dozens of heating stations with its own local network, and random changes occurring in any of the elements are more noticeable in it. In this case, since each thermal station is autonomous, it can be a normal and efficient source of heat due to the peculiarities of energy control based on its own hydrodynamic properties. At the same time, it is problematic to take into account all influencing factors in specific territories as well as to collect geographically distributed information. In this regard, the development of analysis methods and the prediction of the operation of thermal centers based on the modern theory of analysis and modeling of dynamic processes that are convenient for implementation and adequately describe real physical situations are relevant. In this case, the most important task of the analysis of heat consumption is to determine the patterns of change in the studied phenomenon, which are formed under the influence of a set of reasons that act on it constantly for a long time. These reasons are sometimes completely random, which complicates the study process and can lead to an incorrect conclusion. To get out of this situation, it is necessary to use a fuzzy logic apparatus, which reduces the influence of random in-row changes by introducing linguistic terms. With this in mind, the goal of the article is to build a prognostic model based on the theory of fuzzy logic parameters, which may be key to the development of the energy-management process based on archived data from thermal centers. A method for constructing a forecast model of a fuzzy approach for a time series with in-row multiplicative changes and the possibility of using it for operational dispatch analysis of heat-supply systems are presented. The simulation is based on the calculation of predicted values for the membership function built on the statistical indicators of row-by-row changes. The proposed method is applied to the time series of dynamic parameters in heat-supply systems. Based on the performed experiments and computational experiments, it was found that the average relative error between the predicted and actual indicators was 3.5% for the temperature in the supply and 4.7% in the return pipe of the system.

中文翻译：

---

供热系统运行调度分析中的模糊预测模型

摘要

大城市的供热网络是一个复杂的水力系统，由数十个供热站组成，并具有自己的本地网络，其中任何元素发生的随机变化在其中尤为明显。在这种情况下，由于每个热站都是独立的，因此由于其自身的流体力学特性而具有能源控制的特殊性，因此它可以成为常规高效的热源。同时，要考虑到特定地区的所有影响因素以及收集地理分布的信息是有问题的。在这方面，基于现代过程分析和建模理论的热分析中心方法的开发和热中心运行的预测非常重要，这些理论便于实施并充分描述实际物理情况。在这种情况下，热量消耗分析的最重要任务是确定所研究现象的变化模式，这些变化模式是在长期持续对其起作用的一系列原因的影响下形成的。这些原因有时是完全随机的，这会使研究过程变得复杂，并可能导致错误的结论。为了摆脱这种情况，必须使用模糊逻辑设备，该设备通过引入语言术语来减少行内随机变化的影响。考虑到这一点，本文的目标是基于模糊逻辑参数理论构建一个预测模型，这可能是基于热中心归档数据开发能源管理过程的关键。提出了一种建立行内乘法变化的时间序列模糊方法的预测模型的方法，以及将其用于供热系统的运行调度分析的可能性。该模拟基于对隶属函数的预测值的计算，该预测函数建立在逐行变化的统计指标之上。该方法适用于供热系统中动态参数的时间序列。根据进行的实验和计算实验，发现对于系统中的温度，预测指标与实际指标之间的平均相对误差为3.5％，而在回流管中则为4.7％。该模拟基于对隶属函数的预测值的计算，该函数基于逐行变化的统计指标建立。该方法适用于供热系统中动态参数的时间序列。根据进行的实验和计算实验，发现对于系统中的温度，预测指标和实际指标之间的平均相对误差为3.5％，而在回流管中的平均相对误差为4.7％。该模拟基于对隶属函数的预测值的计算，该预测函数建立在逐行变化的统计指标之上。该方法适用于供热系统中动态参数的时间序列。根据进行的实验和计算实验，发现对于系统中的温度，预测指标与实际指标之间的平均相对误差为3.5％，而在回流管中则为4.7％。