论文部分内容阅读
针对国内外水电站实际调度水平与优化调度理论间差距普遍较大的现状,基于确定性优化调度成果,运用支持向量机理论建立了水库水电站实际运行的中长期调度计划,以提高水电站实际运行调度水平。结合决策变量与相关因子筛选、训练样本与预测样本划分、模型参数优选等技术处理,将调度计划应用于金沙江—长江中游水电站群中的4座月调节性能以上水电站,预测结果表现出了良好的拟合精度。进一步模拟调度计划,指导梯级水电站群长系列调度运行,结果表明,模拟运行较好地继承了确定性优化调度在发电量、发电保证率及发电过程方面的效益,证实了该方法的可行性和有效性,为梯级水电站群实际优化运行提供了有力的技术支撑和决策支撑。
In view of the current situation that the disparity between actual dispatching level of hydropower station and optimal dispatching theory is generally large, based on deterministic optimal dispatching result, a medium- and long-term dispatching plan of reservoir hydropower station is established by using support vector machine theory to improve the actual operation dispatching level of hydropower station . Combined with the selection of decision variables and relevant factors, the training samples and prediction samples, and the optimization of model parameters, the scheduling scheme is applied to the hydropower station above the 4-month regulation level in the Jinshajiang-Changjiang middle reaches hydroelectric power station group. The prediction results show good results Fitting accuracy. The results show that the simulated operation inherits the benefit of the deterministic optimal dispatch in terms of power generation, power generation guarantee rate and power generation process, which confirms the feasibility of the method and the feasibility of the method. Validity, provide a strong technical support and decision support for the actual optimization operation of cascade hydropower stations.