基于响应曲面模型的泵站进水池参数优化方法研究

Research on optimization method of pump sump parameters based on response surface model

泵站进水池的结构参数,特别是吸水喇叭管参数取值对泵站整体流态影响很大。本文基于响应曲面模型和CFD数值模拟,以垂直布置的吸水喇叭管悬空高、后壁距和淹没深度为设计变量,以进水池压力场、速度场和涡量场评价指标的加权函数为目标函数,进行2因素和3因素的优化设计分析。研究结果表明,变量之间对进水池流态的交互作用较大,淹没深度对最优悬空高、后壁距的影响不可忽略,且淹没深度与悬空高的交互作用(P=0.0019)较其与后壁距的交互作用(P=0.0696)大。针对本文垂直布置型式的吸水喇叭管,推荐最优参数组合是悬空高0.77 D,后壁距0.37 D,淹没深度2.19 D。在该组合下实际值与预测值误差不大于4.82%,与泵站设计规范推荐值相比,其计算域水力损失降低0.98%,流速分布均匀度提高5.92%,机组纵剖面涡量分布特征值降低3.1倍,大大改善了进水池的流态。这表明所建立的二次多项式响应面模型能够较精确地表示设计变量与目标函数之间的关系,基于响应曲面模型的优化设计方法可有效用于泵站进水流场的流态优化。

The parameters of pump sump, especially the parameters of bell-mouth are significantly impor- tant for the flow pattern of pumping station. Thus, the floor clearance, back wall clearance and submer- gence of bell-mouth are taken as the design variables, and the weighting function composed of evaluation indexes for pressure, velocity and vortex is taken as objective function. Therefore, optimization design and analyses of two and three factors have been done on the basis of response surface model and CFD simula- tion. The results show that the interactions between variables and the flow pattern are significant, and the impact of submerged depth on floor clearance and back wall clearance cannot be ignored. Furthermore, the interaction （P=0.0019） between submergence and floor clearance is more significant than that （P=0.0696） of submergence and back wall clearance. The optimized pretreatment parameters are as follows： the floor clearance is 0.77D, and D indicates the bell-mouth diameter; the back wall clearance is 0.37D, and the submergence is 2.19D. Under the optimized parameters, the error between the actual value and the predict value of objective function is less than 4.82%. Comparing with the recommended parameters of Design code for pumping station, the hydraulic efficiency of computational domain and the uniformity of axial velocity distribution are 0.98% lower and 5.92% higher, respectively; and the vortex distribution value of pump sump profile is reduced by about three times, which greatly improved the flow pattern of the pumping sta- tion. So it can be concluded that the relationship between the design variables and the objective functions can be accurately presented by the two order polynomial response surface model, and the optimization meth- od based on response surface model can be effectively used for the optimization of flow in pumping station.