梯级水电站近边界运行全景模糊风险分析及多属性方案决策

This project takes the cascade hydropower stations of Yalong River in our country as the research object. We first analyze the fuzzy distribution characteristic of runoff forecasting error, and derive its fuzzy membership function and credibility distribution function, so that to establish a quantitative mathematical description method for the runoff forecasting error in fuzzy environment, and to study and discuss the effective coupling mode of the fuzziness of forecasting error in the risk analysis of short-term power generation operation of cascade hydropower stations. And then, aim at the special near boundary operation area of hydropower station, we set up the runoff forecasting error considered panoramic fuzzy risk analysis model for the near boundary operation of cascade hydropower stations, and put forward an efficient solving method based on fuzzy simulation technology and multi-level nested parallel mode, and build a multidimensional panoramic fuzzy risk space named "near boundary water level combination - comprehensive operation risk" by solving the model, so that to reveal the influence mechanism and law of the fuzziness of runoff forecasting error for the power generation operation risk of cascade hydropower stations. On this basis, we propose the risk and benefit indicators that affect the safety and economy of the power generation plan under the influence of uncertain runoff, and then establish the synergy evaluation index system of risk and benefit for the power generation plan, and put forward a multi-attribute evaluation and decision making model for the power generation plan, and then the optimal power generation plan for cascade hydropower stations under different near boundary conditions can be proposed by this model. So, this project can improve the utilization level of water resource, and will provide a big development and improvement for the theories and methods used in the risk control and decision making of short-term power generation operation of cascade hydropower stations.

项目以雅砻江梯级水电站为研究对象，首先探讨径流预报误差的模糊分布特征，推导其模糊隶属度函数及可信性分布函数，提出模糊环境下径流预报误差的定量数学描述方法，并探究预报误差模糊性在水电站发电调度风险分析中的耦合方式；其次针对水电站近边界这一特殊运行区域，建立考虑径流预报误差模糊性的梯级水电站近边界运行全景模糊风险分析模型，提出基于模糊模拟技术和多层级嵌套并行模式的模型高效求解方法，构建“近边界水位组合-综合调度风险”全景模糊风险空间，以揭示径流预报误差模糊性对梯级水电站发电调度风险的影响机理与规律；在此基础上，提出径流不确定情形下影响发电计划方案安全性和经济性的风险与效益指标，构建发电计划方案风险~效益协同评价指标体系，并建立多属性方案评价决策模型，以此提出梯级水电站在不同近边界运行工况下的最佳发电计划方案，提高流域水资源利用水平，发展和完善梯级水电站短期发电调度风险控制与决策的理论方法体系。

水电站短期发电计划编制采用预报径流进行，但受预报精度限制，预报值与实际值往往存在偏差，当电站位于近边界水位运行时，该偏差就可能带来弃水或出力不足问题。鉴于此，为定量分析径流预报误差的模糊特性及其对水电站短期发电调度的影响，项目以雅砻江流域锦西水电站等为研究对象，主要开展了以下两方面的研究：.（1）首先探讨了径流预报误差的模糊分布特征，推导了其模糊隶属度函数及可信性分布函数，并以多种贴近度准则、多种理论模糊分布函数及遍历收索的参数优化方法为基础，建立了一种基于贴近度最优准则的模糊隶属函数优选模型，实现了模糊环境下径流预报误差模糊量化。以锦西电站为背景进行实例研究，对不同来流量级下径流预报误差模糊集的模糊隶属函数进行了推求和优选。结果表明，相比柯西分布和正态分布，优化后的尖Γ分布能够更好的拟合实际点据情况，其平均贴近度可达到0.979，很好地实现了锦西电站不同来流量级下径流预报误差模糊分布规律的精确数理表达。.（2）为准确评估因径流预报误差所带来的水电站近边界运行时的弃水或出力不足风险，项目基于可行性理论实现了水文预报误差与水电站短期调度风险的耦合，创新性的建立了用于评估水电站近边界运行风险情况的全景模糊风险分析模型，提出了基于模糊模拟技术和多层级嵌套并行模式的模型高效求解方法。以锦西电站为例进行实例研究，构建了“近边界水位组合-综合调度风险”全景模糊风险空间，揭示了径流预报误差模糊性对水电站发电调度风险的影响机理与规律，得出了该电站在不同近边界运行时的高风险区域及建议运行区域，可为水电站在不同近边界运行工况下的最佳发电计划方案制定提供决策支持，具有重要的指导意义。