【作者】 杨佳；

【导师】 丁晓群；

【作者基本信息】 河海大学 ， 电力系统及其自动化， 2006， 硕士

【摘要】 在电力网损中变压器的损耗占有相当大的比例,故实现变压器经济运行是降低网损、节约能源的有效手段之一。但综合国内外的研究现状,如何有效的实施变压器经济运行还没有得到很好解决。本文着重研究了在地区电网现有的SCADA系统和设备上实现在线变压器经济运行的解决方法,主要做了以下四方面的工作: 1.分析推导了变压器经济运行的一些基本问题,主要对变压器的经济负载系数的计算、运行区间的划分、变压器间技术优劣的判定以及并列和分列运行模式下的经济运行方式的确定进行了分析和讨论。 2.通过变电站监控系统得到SCADA实时数据,采用最小正交二乘法,对变压器损耗曲线进行拟合计算,得到变压器实际参数和功率损耗特性,并以此作为变压器经济运行的基础数据。 3.针对负荷波动较大的变电所,提出了一种基于神经网络的短期负荷预测方法。并以负荷预测的结果作为变压器经济运行实时负荷数据的参照。 4.用时段控制法和模糊决策法均能最终确定变电站变压器运行方式改变的时机。根据仿真结果所示,两种方法都效果良好,但模糊决策法更加易于实际工程应用。更多还原

【Abstract】 The loss of power transformers accounts for major proportion in the loss of the power system, so one of the effective measure to reduce loss and save energy is to realize economic operation of the transformer. But from the present status in this field, how to effectively realize economic operation of the transformer has not been settled very well. This dissertation chiefly studies how to effectively realize economic operation of the transformer based on the existing SCADA system and equipments of the district power network. The principal work is as following:Firstly, some basic issues of economic operation of the transformer are analyzed, focusing on calculating the economic load coefficient, carving up the running area of the transformer, determining the quality between transformers, and confirming economic operation mode of transformers in the parallel or divided running pattern.Secondly, a method based on SCADA to revise the parameter of transformers is proposed, in which the data is obtained from the substation SCADA system and is processed by least orthogonal square method, then the fitting calculation of the curve is performed, finally the actual parameters of the transformer can be derived and it can be used as the basic data for the economic operation of the transformer.Thirdly, a short-term load forecasting method based on ANN (artificial neural network) is proposed, as the load fluctuates frequently in some substations. And the real-time data is according to the result of the short-term load forecasting.At last, the optimal switch-on/off time can be determined eventually by the proposed both Time-interval Control Method and Fuzzy Decision. The result of simulation shows that the two method are also effective. But the Fuzzy Decision is easy to apply in practice.更多还原