基于关联规则的发电关键要素研究

水电站发电能力评估是水电站发电调度中的重要内容。从水电站历史运行数据中寻找满足特定电量目标的水位、水量关联规则,对于指导水电站的运行具有重要的参考价值。为此,设计了将发电调度数据转化为可用于数据挖掘格式的方法,对相关变量取值范围进行了率定,其中包括库水位的取值从死水位到校核洪水位、发电量取值从最小值到装机容量以及入库水量和区间来水量按频率曲线取值,将其各变量划分为10个区间,并给各区间赋予了相应的标签,使之可应用于数据挖掘。在此基础上,利用FP-tree算法,计算了不同月份和电量目标条件下乌江梯级某两座水电站之间的水位、入库水量等相关因素的关联规则,并给出了要满足特定电量目标的最低水位、入库水量、区间水量之间的耦合关系。

Study on key factors of power generation based on association rules

The assessment of power generation ability is an important aspect in the regulation of hydropower plant. The correlation between targeted power generation quantity and the water level, water quantity obtained from historical data has great referential value to the regulation of hydropower plant. In this paper, the methods for format conversion were proposed, which made the data of hydropower plant generation regulation be available to data mining through the way that calibrates the range of related parameters. In detail, the range of water level was from dead level to the maximum flood level, the range of power generation was from the minimal to the installed capacity, and the range of water inflow and interval flow were determined by the frequency curve. All the parameter ranges were divided into 10 intervals, and each of them was labeled with certain name. Based on this, using FP-tree algorithm and taking two hydropower plants on Wujiang River as example, the association rules of water level and water quantity were calculated in the conditions of various months and power generation aims. The coupled relations of lowest levels, inflow and interval water quantity were discussed to meet the given targeted power generation.