改进 PSO-GA 混合优化下高温合金加工性能预测模型

Prediction Model of Superalloy Machinability Under Improved PSO-GA Joint Optimization

摘要

镍基高温合金由于其硬度高、热稳定性好,是航空发动机制造业中的关键材料,但同时也是难加工材料,因此,通过建立预测模型提高其加工性能对于航空制造业具有重要意义。研究采用了正交实验法来确定外圆磨削GH3044阶梯轴的工艺参数,包括砂轮转速、工件转速、往复速度和磨削深度,并通过拉格朗日插值法和归一化对采集数据进行预处理。在此基础上,采用改进的粒子群-遗传混合算法对反向传播神经网络模型进行优化以期获得更精确的预测模型,优化内容包括隐藏层节点数、激活函数、初始权值和偏置。经留一法交叉验证后的结果表明,优化后的反向传播(backpropagation,BP)神经网络模型对工件的主要加工性能表面粗糙度R_a和去除效率e的相关系数分别达到了0.95938和0.90592。研究得出的预测模型有望为磨削镍基高温合金的工艺提供重要指导,有助于保持高精度加工,提高生产率,同时降低生产成本。

Abstract

Nickel-based superalloys are key materials in the aero-engine manufacturing industry due to their high hardness and good thermal stability, however, they are also difficult to machine. Therefore, improving its machinability by establishing a predictive model is of considerable significance for the aerospace manufacturing industry. The study employed orthogonal experimental method to determine the process parameters of cylindrical grinding GH3044 stepped shaft, including grinding wheel speed, workpiece speed, reciprocating speed and grinding depth. On this basis, an improved particle swarm optimization-genetic algorithm (PSO-GA) was used to optimize the back-propagation neural network model (BPNN) yielding a more accurate prediction model, which includes the number of nodes in the hidden layer, activation function, initial weights and bias. The optimized BPNN model achieves correlation coefficients of 0.95938 for surface roughness R_a and 0.90592 for removal efficiency e, which are the main machinability of the workpieces, and the optimized BPNN model is expected to provide important guidance for the process of grinding nickel-based superalloys, which can help to maintain high-precision machining, improve productivity, and reduce the cost of production.

关键词

镍基高温合金 / 磨削加工 / 数据预处理 / 预测模型 / 优化算法 / 表面质量 / 去除效率

Key words

nickel-based superalloys / grinding / data preprocessing / predictive model / optimization algorithm / surface quality / removal efficiency

参考文献

[1] 胡壮麒, 刘丽荣, 金涛, 等 . 镍基单晶高温合金的发展[J]. 航空发动机, 2005, 31(3): 1-7.

[2] Hu Z Q, Liu L R, Jin T, et al. Development of the Ni—base single crystal superalloys[J]. Aeroengine, 2005, 31(3): 1-7. (in Chinese)

[3] 王昊, 荆鹏, 李广忠 . 烧结温度对多孔Inconel 625合金性能的影响[J]. 稀有金属材料与工程, 2021, 50(10): 3693-3697.

[4] Wang H, Jing P, Li G Z . Effect of sintering temperature on properties of porous Inconel 625 alloy[J]. Rare Metal Materials and Engineering, 2021, 50(10): 3693-3697. (in Chinese)

[5] 蔡明, 朱涛, 高兴军, 等 . 镍基高温合金磨削加工技术研究现状[J]. 辽宁石油化工大学学报, 2023, 43(3): 60-68.

[6] Cai M, Zhu T, Gao X J, et al. Research status of grinding technology of nickel—based superalloy[J]. Journal of Liaoning Petrochemical University, 2023, 43(3): 60-68. (in Chinese)

[7] 王冬冬, 王波 . GH4169抗疲劳表面砂带磨削工艺参数优化[J]. 航空制造技术, 2023, 66(21): 125-131.

[8] Wang D D, Wang B . Optimization of process parameters for GH4169 anti—fatigue surface abrasive belt grinding[J]. Aeronautical Manufacturing Technology, 2023, 66(21): 125-131. (in Chinese)

[9] Ji T, Huang S L, Ren B M, et al. Study on the characteristics of high temperature alloy surface profile grinding process[J]. Materials Research Express, 2022, 9(6): 066506.

[10] 田凤杰, 吕冲 . 基于BP神经网络的机器人砂带磨削表面粗糙度研究[J]. 工具技术, 2018, 52(9): 100-103.

[11] Tian F J, Lü C . Research on surface roughness of robotic abrasive belt grinding based on BP neural network[J]. Tool Engineering, 2018, 52(9): 100-103. (in Chinese)

[12] 秦娜, 刘凡, 刘亚龙, 等 . 基于遗传算法优化BP神经网络的钛合金旋转超声磨削力预测[J]. 中国科技论文, 2017, 12(10): 1128-1131, 1156.

[13] Qin N, Liu F, Liu Y L, et al. Prediction of grinding force in rotary ultrasonic grinding of titanium alloy based on BP neural network optimized by genetic algorithm[J]. China Sciencepaper, 2017, 12(10): 1128-1131, 1156. (in Chinese)

[14] Deng Z H, Zhang X H, Liu W, et al. A hybrid model using genetic algorithm and neural network for process parameters optimization in NC camshaft grinding[J]. The International Journal of Advanced Manufacturing Technology, 2009, 45(9/10): 859-866.

[15] 孙杨, 于联周, 巩亚东 . 单晶材料微磨削表面粗糙度模型预测与优化[J]. 中国工程机械学报, 2022, 20(4): 300-305.

[16] Sun Y, Yu L Z, Gong Y D . Predictive and optimal model of surface roughness in micro—grinding single crystal material[J]. Chinese Journal of Construction Machinery, 2022, 20(4): 300-305. (in Chinese)

[17] 曾权, 李鑫, 王克鲁, 等 . 基于GA—BP和PSO—BP神经网络的SLM GH3625高温合金残余应力预测研究[J]. 塑性工程学报, 2024, 31(3): 193-199.

[18] Zeng Q, Li X, Wang K L, et al. Study on residual stress prediction of SLM GH3625 high temperature alloy based on GA—BP and PSO—BP neural networks[J]. Journal of Plasticity Engineering, 2024, 31(3): 193-199. (in Chinese)

[19] Zhang B C, Chen S Y, Gao S Y, et al. Combination balance correction of grinding disk based on improved quantum genetic algorithm[J]. IEEE Transactions on Instrumentation and Measurement, 2023, 72: 1000112.

[20] Chen Z Y, Li X K, Zhu Z Y, et al. The optimization of accuracy and efficiency for multistage precision grinding process with an improved particle swarm optimization algorithm[J]. International Journal of Advanced Robotic Systems, 2020, 17: 1729881419893508.

[21] 盛仲飙 . BP神经网络原理及MATLAB仿真[J]. 渭南师范学院学报, 2008, 23(5): 65-67.

[22] Sheng Z B . BP neural network principles and MATLAB simulation[J]. Journal of Weinan Normal University, 2008, 23(5): 65-67. (in Chinese)

[23] 刘志峰, 潘丹, 王建华, 等 . PSO—BP神经网络在MBR工艺中的膜污染预测[J]. 北京工业大学学报, 2012, 38(1): 126-131.

[24] Liu Z F, Pan D, Wang J H, et al. Flux prediction of MBR based on PSO—BP neural network[J]. Journal of Beijing University of Technology, 2012, 38(1): 126-131. (in Chinese)

[25] 陈际, 袁守利, 刘志恩 . 基于BAS改进PSO算法对PEMFC温度的控制[J]. 太阳能学报, 2023, 44(5): 67-73.

[26] Chen J, Yuan S L, Liu Z E . Temperature control of PEMFC based on BAS improved PSO algorithm[J]. Acta Energiae Solaris Sinica, 2023, 44(5): 67-73. (in Chinese)

[27] 吕翀, 庞永杰, 张磊 . 基于MPSO算法的水下机器人PD控制器[J]. 北京工业大学学报, 2010, 36(6): 728-734.

[28] Lü C, Pang Y J, Zhang L . AUV PD controller based on MPSO algorithm[J]. Journal of Beijing University of Technology, 2010, 36(6): 728-734. (in Chinese)

[29] 赵宏岩, 邬昌兴 . 基于改进粒子群算法的计算机考试自动组卷方法[J]. 佳木斯大学学报(自然科学版), 2024, 42(10): 5-9.

[30] Zhao H Y, Wu C X . Automatic grouping method for computerized exams based on improved particle swarm optimization[J]. Journal of Jiamusi University (Natural Science Edition), 2024, 42(10): 5-9. (in Chinese)

[31] 李文华, 牛国波, 刘羽佳 . 基于遗传算法优化BP神经网络的液压系统故障诊断[J]. 机床与液压, 2023, 51(8): 159-164.

[32] Li W H, Niu G B, Liu Y J . Hydraulic system fault diagnosis based on genetic algorithm optimized BP neural network[J]. Machine Tool & Hydraulics, 2023, 51(8): 159-164. (in Chinese)

[33] 姜菲菲, 赵凤霞, 牛森涛, 等 . 采用改进遗传算法的动力电池成组技术研究[J]. 机械设计与制造, 2024(2): 241-245.

[34] Jiang F F, Zhao F X, Niu S T, et al. Research on grouping of power battery based on improved genetic algorithm[J]. Machinery Design & Manufacture, 2024(2): 241-245. (in Chinese)

[35] 叶帅, 余江鸿, 姚齐水, 等 . 基于粒子群—遗传混合算法的深沟球轴承优化设计[J]. 湖南工业大学学报, 2024, 38(1): 32-39.

[36] Ye S, Yu J H, Yao Q S, et al. An optimized design of deep groove ball bearings based on particle swarm—genetic hybrid algorithm[J]. Journal of Hunan University of Technology, 2024, 38(1): 32-39. (in Chinese)

[37] Chen X D, Ge X X, Sun R F, et al. A SVM based demand response capacity prediction model considering internal factors under composite program[J]. Energy, 2024, 300: 131460.

[38] 杨少瑜, 黄国栋, 林星宇, 等 . 基于拉格朗日插值法的概率建模方法及其在概率潮流分析中的应用[J]. 现代电力, 2021, 38(4): 378-385.

[39] Yang S Y, Huang G D, Lin X Y, et al. A Lagrange interpolation based probabilistic modeling method and its application in probabilistic power flow analysis[J]. Modern Electric Power, 2021, 38(4): 378-385. (in Chinese)

[40] Chipanshi A, Zhang Y S, Kouadio L, et al. Evaluation of the Integrated Canadian Crop Yield Forecaster (ICCYF) model for in—season prediction of crop yield across the Canadian agricultural landscape[J]. Agricultural and Forest Meteorology, 2015, 206: 137-150.

[41] 杨柳, 王钰 . 泛化误差的各种交叉验证估计方法综述[J]. 计算机应用研究, 2015, 32(5): 1287-1290, 1297.

[42] Yang L, Wang Y . Survey for various cross—validation estimators of generalization error[J]. Application Research of Computers, 2015, 32(5): 1287—1290, 1297. (in Chinese)

基金资助

国家科技重大专项资助项目(2021-0171-1-1)