基于WPSO-BP和L-MBWO的多翼离心风机优化研究  

Optimization of multi-blade centrifugal fan based on WPSO-BP and L-MBWO

作　　者：徐韧 李君宇 周明[2,3] 刘林波[2,3] 张志富 黄其柏 XU Ren;LI Junyu;ZHOU Ming;LIU Linbo;ZHANG Zhifu;HUANG Qibai(School of Mechanical Science and Engineering,Huazhong University of Science and Technology,Wuhan 430074,China;Valeo-Huazhong University of Science and Technology Joint Laboratory of Vibration and Noise,Wuhan 430074,China;Valeo Automotive Air Conditioning Hubei Co.,Ltd.,Jingzhou 430074,China;School of Mechanical and Electrical Engineering,Hainan University,Haikou 570228,China)

机构地区：[1]华中科技大学机械科学与工程学院,湖北武汉430074 [2]法雷奥-华中科技大学振动与噪声联合实验室,湖北武汉430074 [3]法雷奥汽车空调湖北有限公司,湖北荆州430074 [4]海南大学机电工程学院,海南海口570228

出　　处：《机电工程》2024年第10期1833-1843,共11页Journal of Mechanical & Electrical Engineering

基　　金：广西科技重大专项(桂科AA22068060-6,桂科AA23062073-3)。

摘　　要：针对多翼离心风机气动性能、噪声情况难以同时改进的问题,提出了一种基于变权重粒子群优化算法的反向传播神经网络风机性能预测模型(WPSO-BP),以及一种基于逻辑混沌初始化的多目标白鲸优化算法(L-MBWO),并将二者应用于多翼离心风机的优化设计中。首先,选取了叶片进出口角、倾斜蜗舌的最大蜗舌半径、叶片切除角度作为设计变量,把风机的全压、效率、声压级作为优化目标;然后,构建了WPSO-BP预测模型,以反映设计变量与优化目标之间的关系,定量分析对比了该模型与BP神经网络预测模型,预测值用于风机的性能优化;接着,将逻辑混沌初始化引入到白鲸优化算法(BWO),基于第三代非支配排序遗传算法(NSGA-Ⅲ)构建了L-MBWO优化算法;最后,在实验验证仿真可靠的前提下,将提出的预测模型和优化算法应用于风机优化,并对优化效果进行了综合分析。研究结果表明:优化后的风机全压增加了34.79 Pa,效率提高了0.67%,噪声降低了1.73 dB,实现了多个优化目标之间的平衡,有效改善了风机的综合性能,为多翼离心风机的优化设计提供了一种新思路。Aiming at the problem that it is difficult to improve the aerodynamic and noise performance of multi-blade centrifugal fans simultaneously,a back-propagation neural network fan performance prediction model based on variable weight particle swarm optimization algorithm(WPSO-BP),and a multi-objective beluga whale optimization algorithm based on logistic chaos initialization(L-MBWO)were proposed,and they were applied to the optimization design of multi-blade centrifugal fan.Firstly,the inlet and outlet angles of the blades,the maximum inclined tongue radius,and the blade cutting angle were selected as design variables,with the total pressure,efficiency,and sound pressure level of the fan as optimization objectives.Then,the WPSO-BP prediction model was constructed to reflect the relationship between design variables and optimization objectives.Quantitative analysis was used to compare the reliability of the WPSO-BP prediction model and the BP neural network prediction model,it indicated that the predicted values could be used to optimize the performance of the fan.Subsequently,the logistic chaos initialization was introduced into the beluga whale optimization(BWO),and L-MBWO optimization algorithm was constructed based on the third-generation non-dominated sorting genetic algorithm(NSGA-III).Finally,the proposed prediction model and optimization algorithm were applied to the optimization of the fan under the premise of experimentally verifying the reliability of the simulation,and the optimization effect was comprehensively analyzed.The research results show that the total pressure of the optimized fan is increased by 34.79 Pa,the efficiency is increased by 0.67%,and the noise is decreased by 1.73 dB,achieving a balance between multiple optimization objectives,effectively improving the comprehensive performance of the fan,and providing a new idea for the optimization design of multi-blade centrifugal fans.

关 键 词：多翼离心风机 变权重 基于变权重粒子群优化算法的反向传播神经网络风机性能预测模型 白鲸优化算法 基于逻辑混沌初始化的多目标白鲸优化算法 预测模型 风机全压 风机效率 风机噪声 

分 类 号：TH432[机械工程—机械制造及自动化] U463.851[机械工程—车辆工程]