论文部分内容阅读
对于日调节或径流式电站而言,调度期初始水位估算的准确与否直接影响其日前计划执行的可行性。考虑到水轮机运转特性曲线(NHQ)方法受计划出力过程影响较大,导致水位估算精度不高,而且其影响因素改进空间有限,故引入1种基于相关因素匹配的模糊聚类分析方法预测电站在调度期的平均耗水率,并利用水量平衡原理计算其初始水位。该方法首先利用已有资料对影响耗水率的所有因素进行权重率定,然后根据所得的最佳权重值通过聚类分析原理预测调度期的平均耗水率,最终得到电站的日初水位。实际算例表明,与NHQ水位估算方法相比,该方法得到的水位估算值满足精度要求,是可行且有效的。
For day regulation or run-of-river power station, the accuracy of the initial water level estimation during the dispatch period directly affects the feasibility of its planned implementation. Considering that the NHQ method is greatly affected by the planned output process, resulting in low estimation precision of water level and limited improvement of its influencing factors, a fuzzy clustering analysis based on correlation factor matching is introduced to predict the power plant The average water consumption during the planning period, and the use of water balance principle to calculate the initial water level. In this method, all the factors affecting the water consumption rate are firstly weighted according to the available data, and then the average water consumption rate of the scheduling period is predicted by the principle of cluster analysis according to the optimal weight value obtained. Finally, the water level at the beginning of the plant is obtained. The actual example shows that compared with the NHQ water level estimation method, the water level estimation obtained by the method meets the accuracy requirements and is feasible and effective.