变化环境下西南河流源区来水-供水-发电-环境互馈博弈自适应协同调控

By introducing mutlti- disciplinary theories of Synergetics, Self-Adaptive Control, Random Forest and Game Theory into the water reources systems, this research aims to inaugurate several researches on high efficiency water utilization in the River Source Areas in Southwest China with plateau ecology and international rivers. The study is focused in the Lancangjiang Basin: (1) To identify the frangibility and sensitivity of water resource system, and consider the risk of change of runoff in the Lancangjiang Basin. (2) Quantitatively reveal the dynamic mutual feedback actions and game relationships among the nonstationary processes of runoff - water supply – hydropower water use– ecoenvironment water demand under changing environment. (3) Put forward a new concept of a basic equilibrium for dynamic game with ‘nonstationry’ incomplete information, i.e., ‘nonstationry’ perfect Bayesian - Nash equilibrium, which will improve the ‘stationary’ problem (stationary is no longer exist in water resources system under such a changing environment) for traditional dynamic game in water allocation. (4) Develop a water demand prediction model for the highly vulnerable and sensitive water system with mutual feedback and game among mutiple sides and elements of runof and water supply in the Lancangjiang Basin. (5) Develop an adaptive coordinating regulation model for the water resources system with game and mutual feds among runoff - supply - hydropower water use- ecoenvironment water demand in the Lancangjiang Basin based on theories of Synergetics and Self-Adaptive Control. (6) Scheme the countermeasures of coordinately sustainable utilizating water resources in the Lancangjiang Basin together with the demand for water security for Silk Road Economic Zone.

本研究引入协同学、自适应控制、随机森林算法、博弈论等交叉学科理论技术，识别变化环境下西南河流源区水资源系统脆弱度和敏感度，量化解析高强度变化环境下“非一致性”来水与供水以及发电用水、生态需水之间的动态互馈作用与博弈关系，提出“‘非一致性’不完全信息的动态博弈基本均衡——‘非一致性’精炼贝叶斯—纳什均衡”概念，改进传统的水资源配置中“平稳”序列的不完全信息的动态博弈问题；提出西南河流源区高度敏感脆弱与供需水多要素博弈的水资源系统的多边多元需水预测理论方法；构建基于自适应控制理论、协同学原理的西南河流源区来水-供水-发电-环境多元多边互馈博弈的自适应协同调控模型，提出“丝绸之路”背景下西南河流源区水资源适应性利用方案和对策，实现西南河流源区脆弱性敏感性水资源系统多边多元供需水互馈博弈的自适应可持续协同调控。选定澜沧江流域为研究区域。

本项目采用澜沧江流域上游树轮年表设计了一种多元线性回归法对流域449年（公元1557-2005）的长系列年径流进行重构的方案，识别了变化环境下流域径流长期演变趋势、周期；提出了流域控制站非一致性来水过程基于协变量的极端来水频率计算方法；针对澜沧江流域实测数据不足的问题，开发了一种结合实测降雨数据，并考虑气候预测系统再分析（CFSR）降雨数据空间异质性，来修正日CFSR降雨数据和澜沧江径流模拟的方法；提出了澜沧江流域基于Budyko假设的傅抱璞公式的水资源敏感性计算方法，揭示了流域水资源系统脆弱性；研制了需水预测的两阶段校正预测技术方法，合理处理了多变多元需水博弈下的预测不确定性；探究并建立了澜沧江流域控制站来水—发电—生态—航运—出境水等多要素互馈博弈关系的数学表达，分别给出了丰、平、枯水年流域来水、发电用水、河道内生态用水、航运用水和出境水博弈竞争的最优纳什均衡策略典型过程，景洪以下河道枯季日流量对出境和航运需水的保证率分别由常规调度的78.4%、40.3%提高到博弈均衡调度的93.7%和48.8%；建立了基于供需水互馈博弈“假设-模拟-反馈-调整”反馈回路的自校正水资源优化配置方法与模式，解决了水资源配置过程中的偏差处理问题，供水保证率、平均缺水率、供水事故发生频次和最长连续缺水时间等4个指标较传统配置改进25%-35%；预测了未来气候预估情景下澜沧江水能资源变化及其开发风险与水电站运行效率，发现气候变化导致澜沧江流域水文情势变异增强，对小湾、漫湾、大朝山、糯扎渡和景洪等梯级电站发电产生显著影响，主要表现在未来水文情势更趋集中化分布，导致发电弃水增加，枯季发电量减少。成果并在珠江流域得到成功应用。发表学术论文94篇，其中SCI论文64篇，ESI高被引论文4篇。主办国际学术会议2次、国内学术会议1次，参加国际会议60人次，授权发明专利1项，获得国家自然科学基金中—阿水资源国际合作项目1项，团队成员中2人晋升副教授，培养博士研究生11名、硕士研究生6名。圆满完成了研究任务。