长江上游水库群复杂多维广义耦合系统调度理论与方法

Aiming to solve several key science problems in scheduling of the upstream cascaded reservoirs of Yangtze River, this project involves analyzing the characteristics of the hydrology dynamical system. By this way, the physical mechanisms of the model predictability will be discussed. Thereafter, it is expected to illustrate chaotic evolution laws of the runoff processes in order to improve the prediction ability of hydrological model. This proposal also attempts to explain the compensatory mechanism of reservoirs storage, and establish the basin-integrated water resources regulation model which meets the water supply needs of the middle and lower reaches. It is expected that the flood control and generation scheduling model with random multivariable coupling will be proposed, by exploring the basic physical background in the adaptive dynamic nested scheduling. Different scheduling objectives which are to be studied include identical and opposing relationship in complex operation conditions. Afterwards, research work attempts to find a dimension reduction method as well as the optimum solution set scale which can reflect the characteristics of non-inferior frontier. In this way, the system dynamics method of multi-objective balance scheduling in a multi-constraint situation will be established. This research effort will recognize the action mechanism of cascaded reservoirs operation on basin ecological system, and a novel physical process will be observed in compensatory regulation of ecological restoration and habitat improvement. Through revealing the response pattern between scheduling scheme of cascaded reservoirs operation and the variation of key risk factor at different confidence levels, the theory bottleneck of multi-attribute risk decision making theory on joint operation of cascaded reservoirs could be overcome. In addition, the mathematical description method for complex decision problems will be proposed on the multi-dimensional generalized coupling hydropower energy system, and the countermeasures theory and method for supporting cooperative scheduling under the meaning of Bayesian equilibrium will be developed.

通过流域水文动力系统特性分析，认识流域径流的可预报性，阐明复杂径流过程混沌演化机理，提高径流预报能力；研究库群多维时空尺度库容补偿机制，建立满足中下游供水需求的流域一体化水资源调蓄模型；探究水库群自适应动态嵌套调度的基本物理背景，提出多随机变量耦合的防洪发电调度模型；解析复杂运行条件下库群不同调度目标的同一和对立关系，推求反映非劣前沿特性的最优方案集规模和目标函数降维方法，建立多重约束条件下多目标均衡调度的系统动力学方法；认识梯级调度对流域生态系统的作用机理，在流域生态修复和生境改善的补偿调控方面揭示新的物理过程；揭示不同置信水平下联合调度方案对关键风险因子变化的响应规律，突破库群联合优化调度多属性风险决策建模的理论瓶颈；提出水电能源多维广义耦合系统复杂决策问题的数学描述方法，发展贝叶斯均衡意义下支持合作调度的对策决策理论与方法。解决长江上游复杂水库群联合优化运行存在的若干关键科学问题。

我国已经形成世界规模最大的长江梯级水库群，承担了防洪、发电、供水、生态和航运等复杂调度任务，其运行调度涉及的领域是空前的，难度史无前例。为此，紧密围绕国家重大需求与现有基础理论间的突出矛盾开展研究，提出水电能源复杂多维广义耦合系统新概念，创建了一套长江上游水库群复杂多维广义耦合系统优化运行的先进理论与方法体系。首先，通过水文动力系统特性分析，揭示了复杂径流过程的混沌现象及其演化机理，阐明了流域径流的可预报性，构建了大气-水文-陆面多源信息驱动的径流非线性高精度预报体系；提出了库群多维时空尺度库容补偿机制和水位消落控制模式，建立了满足中下游供需平衡的流域水资源一体化调蓄模型；探明了长江流域洪水组合遭遇规律，提出了防洪库容优化分配建模方法，建立了库群分层多时空嵌套精细化调度与跨区多电网联合调峰优化调度模型，形成了多随机变量耦合的防洪发电调度模型体系；在此基础上，提出了反映非劣前沿特性的最优方案集规模和目标函数降维方法，建立了多重约束条件下多目标均衡调度的系统动力学方法；洞悉了梯级调度对流域生态系统的作用机理，构建了水库群调度与水动力学联合驱动的生态补偿调度模型，揭示了生境改善与调控新的物理过程；进一步，研究并揭示了不同置信水平下联合调度方案对关键风险因子变化的响应规律，构建了梯级水库群随机调度风险评估模型，提出了梯级水库群多属性风险决策方法，突破了水库群联合优化调度多属性风险决策建模的理论瓶颈；以此为基础，提出了多维广义耦合系统复杂决策问题的数学描述方法，推求了多业主合作调度补偿效益优化分配方法和支持合作调度的动态均衡对策决策模型，建立了水库群调度多目标-多主体-多层次群决策方法体系，发展了贝叶斯均衡意义下支持合作调度的对策决策理论与方法。研究工作解决了长江上游水库群复杂多维广义耦合系统调度中面临的若干核心科学问题和关键技术难题，为我国水能开发和利用提供了科学依据和技术支撑。