基于模拟–优化模式的梯级水库群突发水污染事件应急调度研究

Emergency Dispatching for Cascade Reservoirs of Sudden Water Pollution Accidents Based on the Model of Simulation and Optimization

摘要: 根据突发水污染事件特征及相关影响因素，在水动力模型、水质模型的基础上，建立了基于模拟–优化模式的梯级水库群突发水污染事件应急调度优化模型。以清江中下游为例，采用改进的NSGA-II算法，对水布垭–隔河岩–高坝洲梯级水库群应急调度优化模型的决策变量(应急调度函数中的控制变量)进行多目标优化，得到了Pareto最优解集及相应的Pareto前沿，并探讨了水库群联合调度在应对突发性水污染事件的合理性与有效性。结果表明，在水位、库容允许的情况下，优先调用距离污染源较近的或与被污染河段直接相连的水库，启动应急调度最多可使污染物浓度达标时间减少80.4%。根据优化结果分析了应急调度过程中损失的电能与污染物达标所用时间之间的竞争关系，并定量模拟了突发水污染事件中污染物浓度的沿程变化与随时间变化的过程。

Abstract: According to characteristics and influences of sudden water pollution accidents, an emergency optimal operation model of simulation and optimization was established for cascade reservoirs based on hydrodynamic model and water quality model. The middle and lower reaches of Qingjiang River are selected as the case study. Applying a multi-objective optimization on the decision variables of the model (the control variables of emergency operating rules of QingJiang cascade reservoirs) by using improved NSGA-II algorithm results in the Pareto solutions and Pareto fronts and then the feasibility and effectiveness of emergency dispatching are discussed. The analysis results show that, the reservoirs close to the pollution source or directly connected with the polluted river should be called first in the capacity of reservoir storage. The time of reaching standard can be reduced by up to 80.4% when taking emergency dispatching. According to the optimization results, the competition between the energy lost in the emergency dispatch process and the time reaching standard is analyzed. The distribution of pollutant concentrations and the process of pollutant concentration changes with time are quantitatively simulated.