HSY算法在水质模型参数识别中的应用探讨

参数识别是水质模拟中的技术核心，数据稀缺是水质模型经常面临的问题。采用HSY算法可求得模型参数的分布，而不再是一个单一的最优参数，从而在一定程度上避免了由于“最优”参数失真带来的决策风险，本文以南水北调一期工为例，讨论了在稀疏数据条件下如何应用HSY算法识别CSTR水质模型的参数，并对识别的模型参数进行了检验。模型检验采取了2种方法：分别用动态和稳态计算方法得出的参数进行比较；用模拟结果分布于实测值进行比较。结果表明，HSY算法为稀疏数据条件下进行识别、建立数学模型提供了一条有效的途径。

Approach on Application of HSY Algorithm to Parameter Identification of Water Quality Model

Parameters identification is water quality simulation, but lack of data is an extremely difficult issue. By using HSY algorithm, distribution of modeling parameters can be solved in stead of the single optimistic point, so the risk anamorphosis of parameters can be avoided in some extent. Taking water diversion from South to North project stage I as an example, to discuss the ways for identifying the parameter of a water quality model through a proposed HSY algorithm under spare data. The identified model was then verified and the parameters variability was discussed. Two methods were used to verify the model:lst. Comparison of the parameter values solved with steady and dynamic states respectively, and then compared the results of simulated and actual detected values. The result showed that HSY algorithm is an effective way for identification of model parameter under spare data.