基于PSO的水利蜗壳泵叶轮和导叶匹配设计

Matching design of impeller and diffuser of hydraulic volute pump based on PSO

为改善水利蜗壳离心泵叶轮和导叶的多参数匹配特性,解决性能和几何参数间无精准数学模型的设计难题,提出基于改进粒子群算法的水利蜗壳离心泵水力部件匹配优化方法。首先,采用BladeGen软件对叶轮和导叶进行参数化建模,以三阶贝塞尔曲线分别拟合叶轮和导叶叶片安放角曲线,选取8个控制点(叶轮、导叶各4个)的y坐标值为决策变量,采用CFX软件对不同方案进行定常数值计算得到设计工况效率,并将其作为优化目标;其次,基于叶片泵性能自动优化平台,设定改进粒子群算法种群数和迭代数均为20,自动调动水利蜗壳泵数值计算程序,智能在全局范围内寻优。研究结果表明:在粒子群算法迭代到第13次时,泵效率达到最优,在设计工况下效率提高了3.09%,获得最优参数组合,优化后叶轮和导叶的包角均增大;优化模型内部熵产损失降低,叶轮叶片工作面、叶轮与导叶交界处能量损失明显减小,内部流动更符合叶片型线趋势,流动分离状况得到改善,验证了蜗壳泵智能优化设计方法的合理性。

In order to improve the multi-parameter matching characteristics of the impeller and diffuser of the hydraulic volute centrifugal pump and to solve the design problem of no accurate mathematical model between performance and geometric parameters, a matching optimization method of hydraulic components of the hydraulic volute centrifugal pump based on improved particle swarm algorithm(PSO) was proposed. Firstly, BladeGen software was used to parametrically model the impeller and diffuser, the impeller and diffuser blade angle curves were fitted with third-order Bessel curves, respectively. The y-coordinate values of eight control points(four for each impeller and diffuser) were selected as decision variables. The design operating efficiency was obtained by steady numerical calculations of different solutions using CFX software, and it was used as the optimization target.Secondly, based on the automatic optimization platform of impeller pump performance, the number of populations and iterations of the improved particle swarm algorithm were set at 20, and the numerical calculation program of the hydraulic volute pump was automatically mobilized to find the optimization in the global scope intelligently.The results show that the pump efficiency is optimized at the 13th iteration of the particle swarm algorithm, and the efficiency is increased by 3.09% under the design condition. The internal flow is more in line with the trend of the vane profile, and the flow separation is improved, which verifies the rationality of the intelligent optimization design method of the volute pump.