论文部分内容阅读
弃水是影响梯级水电站经济运行的一个重要因素,针对静态弃水策略的不足,在协调条件的基础上,建立了梯级水电站的动态弃水模型,以此为基础建立调度期内发电量最大、年调节或季调节水库末蓄水量最大、日调节水库末蓄水量偏差平方和最小、总耗水量最小及末级水电站弃水量最小的多目标短期优化调度模型。针对仿电磁学算法原理(electro-magnetism-like mechanism,ELM)简单及收敛迅速的特点,采用嵌入数据包络分析(data envelopment analysis,DEA)的混合仿电磁学算法对多目标优化调度模型进行求解,该算法避免了传统权重系数法的盲目性。对一个8级梯级水电站系统进行仿真分析,结果表明所提出的动态弃水策略可以有效地提高梯级水电站的发电效益,同时也验证了混合仿电磁学算法在求解多目标优化问题时的有效性。
Abandoned water is an important factor affecting the economic operation of cascade hydropower stations. In view of the deficiency of static water abandonment strategy, a dynamic water abandonment model of cascade hydropower stations is established on the basis of coordination conditions. Based on this, The multi-objective short-term optimal scheduling model with annual or seasonal adjustment of maximum storage capacity at the end of the reservoir, minimum square deviation of the daily storage capacity of the daily regulation reservoir, minimum total water consumption, and minimum amount of abandoned water at the final stage hydropower station. Aimed at the simplicity and quick convergence of electro-magnetism-like mechanism (ELM), a hybrid simulation algorithm based on data envelopment analysis (DEA) was used to solve the multi-objective optimization scheduling model This algorithm avoids the blindness of the traditional weight coefficient method. The simulation analysis of an 8 - scale cascade hydropower station shows that the proposed dynamic water abandonment strategy can effectively improve the power generation efficiency of cascaded hydropower stations, and also verifies the effectiveness of the hybrid simulation electromagnetism algorithm in solving multi - objective optimization problems.