粒子群算法在河道水动力模型参数校正中的应用

参数估计一直是河道水动力模型研究的难点之一,在传统的模型参数人为经验率定方法的基础上,提出了基于粒子群算法的模型参数优化校正方法,构建了参数校正优化模型,并将参数优化校正算法与河道水动力模型进行耦合,针对淮河干流和史灌河支流组成的研究区域,采用一维河道洪水演进模型,比较了糙率系数校正方法和传统经验估算法,校正方法得到的河段糙率系数值比人为经验估计值平均大0.01,淮河干流河段糙率略大于史灌河支流河段糙率,采用校正河段糙率系数得到的河道水位过程与实测值拟合更优,特别在主峰段洪水过程模拟精度显著改善,验证了本文所提出的参数优化校正算法的有效性,为复杂河道水动力模型参数的确定提供了一种有效方法。

Application of particle swarm optimization in parameter calibration of channel hydrodynamic model

Parameter estimation has always been one difficulty in channel hydrodynamic model. Based on the traditional method of calibrating model parameters by personal experience, we proposed a method to optimize and correct model parameters based on the Particle Swarm Optimization algorithm, and established an optimization model for parameter correction. Then we coupled the algorithm with the channel hydrodynamic model. We studied the area comprised of the main Huai River and Shiguan River tributary. Using 1D river flood routing model, we compared the roughness coefficient correction method and the traditional empirical estimation method. Results showed that the corrected roughness coefficient was 0.01 larger on average than the experiential roughness coefficient. The roughness in Huai River was slightly larger than the roughness in Shiguan River tributary. The water level hydrograph simulated by the corrected roughness coefficient fit the measured value better than that by the experiential roughness coefficient. Especially , for the main peak period of the flood hydrograph, the simulation accuracy was improved significantly. Thus, the validity of the proposed algorithm was verified. This algorithm provides an effective method for determining the parameters of complex channel hydrodynamic model.