贝叶斯定理构建水库汛限水位动态控制推理模式及其风险

Inference model and its risk analysis of dynamic control of reservoir limit water level based on Bayes theorem

针对现行的汛限水位动态控制推理模式的建模假定合理性证明及风险率计算不完备性问题,以南水北调受水区的白龟山水库资料为例,应用贝叶斯定理证明了现行推理模式的建模假定是成立的,依据完备条件下风险构建新的推理模式较现行的模式结构合理,逻辑性与可操作性强.这一研究成果既有理论意义又有实用价值,可达到既保证防洪安全,又增加洪水资源量的双赢目的,有助于进一步推广应用汛限水位动态控制方法.

Rationality testification of modeling assumptions of inference model and calculation of risk rate under incomplete conditions of dynamic control of the reservoir limit water level aren′t solved.Take Baiguishan Reservoir that is in intake area of south-to-north water diversion as an example,Bayes theorem is applied to testifying that modeling assumptions of inference model of dynamic control of the reservoir limit water level is rational.And a new inference model which is established by risk rate under complete conditions is more logical and feasible than actual inference model.This result can not only ensure the safety of flood control,but also increase the utilization rate of the flood resource.Furthermore,it can provide references with promoting the use of the dynamic control of the reservoir limit water level method.