PEM燃料电池用空压机多参数多目标优化  

Multi-objective and multi-parameter optimization for centrifugal compressor used in PEM fuel cells

作　　者：张良[1,2] 许思传[1,2] 万玉[1,2] 

机构地区：[1]同济大学汽车学院,上海201804 [2]同济大学新能源汽车工程中心,上海201804

出　　处：《电源技术》2016年第1期81-83,93,共4页Chinese Journal of Power Sources

基　　金：国家高技术研究发展计划("863"计划)(2012AA110-501)

摘　　要：为获得适用于燃料电池汽车的高压比、小体积、高效率离心空压机,将均匀设计法、神经网络模型、多目标遗传算法同三维数值模拟相结合,提出了一种离心空压机多目标优化方法。以静压比和效率为优化目标,在额定工况下,对燃料电池高速离心空压机叶轮结构进行优化设计。通过台架实验对三维仿真模型的精确度进行验证,并对优化后的叶轮进行了三维仿真分析。仿真结果表明,优化后叶轮的流动性能得到改善,等熵效率和静压比分别提高4.3%和3.2%,叶轮出口速度更加均匀,验证了该离心空压机多目标优化方法的有效性。In order to design a centrifugal compressor with high pressure ratio, small volume and high efficiency PEMFC vehicles, a novel optimization method was proposed combining the multi-objective genetic algorithm with the uniform design method, the neural network and the three-dimensional numerical simulation. At the rated operating conditions, efficiency and pressure ratio of the compressor were optimized using the proposed method. With the validation of the three-dimensional simulation model by experimental results, the optimized impeller was analyzed by the validated model. Simulation shows that the aerodynamic performances of the optimized impellers are improved. The static pressure ratio and Jsentropic efficiency increased by 4.3% and 3.2%, respectively. In addition, the velocity distribution in the area of impeller outlet became much more uniform. The above results prove that the multi-objective optimization method is effect）re for the design of centritugal compressor.

关 键 词：燃料电池 离心空压机 遗传算法 神经网络 数值模拟 多目标优化 

分 类 号：TM911[电气工程—电力电子与电力传动]