To efficiently predict the forming quality of self-piercing riveted joints, a finite element model of self-piercing riveting for AA5754 aluminum alloys was established, and the effectiveness of the simulation model was verified through experiments. Based on the simulation analysis, 176 sets of effective cross-sectional data of the joints were obtained. By integrating the sparrow search algorithm and the butterfly algorithm, a composite optimization algorithm was constructed. The algorithm's convergence speed and solution quality were improved by employing population initialization and lens reverse learning strategies. Multidirectional learning and Levy flight strategies were introduced to enhance the algorithm's ability to escape local optima, thereby improving the global search capabilities. Research indicates that the prediction results of the established model have a MAPE of less than 10%, a correlation coefficient R2 higher than 0.99, and a mean square error MSE consistently less than 0.001. Therefore, the proposed improved model has high predictive accuracy and robustness.
根据经验公式计算出隐含层神经元个数取值范围为[4,13],学习率设置为0.01,训练目标最小误差为10,迭代次数为1000。训练算法选择结合梯度下降法和高斯-牛顿法的trainlm函数,通过测试集均方根误差确定最佳的隐含层神经元个数。为了消除训练过程随机性对判别结果的影响,对BP神经网络进行了30次训练,最后将训练后的均方根误差(root mean square error, RMSE)取平均值,如图8所示,当隐含层神经元个数为10时,均方根误差最小,因此,建立了结构为5-10-3的初步神经网络预测模型。
2.2 多策略改进复合麻雀搜索算法
BP 神经网络虽然在非线性处理方面具有一定优势,但也存在收敛速度慢、计算过程易陷入局部最优、网络初始权值和阈值的选择对训练结果影响较大等缺点。群智能优化算法则具有出色的全局搜索能力和良好的自适应性,鲁棒性较高,能够有效解决 BP 神经网络存在的问题。本文采用麻雀搜索算法(sparrow search algorithm,SSA)优化BP神经网络。
通常模型预测结果的MAPE 值低于10%属于高精度预测[28],由图14可以观察到,传统 BP 神经网络的RMSE、MAE、MAPE值均较大,模型的鲁棒性较差。而通过GA优化神经网络预测结果的RMSE、MAE、MAPE值进一步降低,证明了对传统 BP 神经网络进行算法优化的必要性。使用MIC_SSA优化BP神经网络,可以看到三个成形截面几何参数的RMSE、MAE、MAPE指标均为最优,大幅提高了预测准确度,结果表明MIC_SSA优化BP神经网络可以有效提高网络模型的预测精度,证明了改进的有效性。
FUPingpo, DINGHua, ZENGXiangrui, et al. Influence of Plate Thickness Combination on Undercut Value of CFRP-Al Self-piercing Riveting Joint[J]. Journal of Plasticity Engineering, 2024, 31(9): 127-134.
LINZhongqin, MAYunwu, XIAYujun, et al. Advances in Forming and Joining Processes of Lightweight High-strength Thin-walled Vehicle Structures[J]. Journal of Mechanical Engineering, 2023, 59(20): 1-17.
LIUYang, ZHUANGWeimin. Research Progress on Clinching Technologies for Metal-polymer and Metal-composite Thin-walled Structures[J]. Materials Reports, 2023, 37(3): 213-224.
[8]
XIEChong, WANGDengfeng, KONGDewen, et al. Material-structure-process-performance Integrated Optimization Method of Steel/Aluminum Self-piercing Riveted Joint[J]. The International Journal of Advanced Manufacturing Technology, 2024, 132(3): 2045-2059.
[9]
MAYunwu, LOUMing, LIYongbing, et al. Effect of Rivet and Die on Self-piercing Rivetability of AA6061-T6 and Mild Steel CR4 of Different Gauges[J]. Journal of Materials Processing Technology, 2018, 251: 282-294.
QINYixin, XINGBaoying, ZHANGHongshen, et al. Forming Process and Mechanical Properties of Self-piercing-through Riveting Joint[J]. Journal of Plasticity Engineering, 2024, 31(5): 48-54.
HUANGZhichao, LILingling, LIHaizhou. Static Strength Analysis of Hollow and Semi-hollow Self-piercing Riveting between Carbon Fiber Composite and 5052 Aluminum Alloy[J]. Journal of Plasticity Engineering, 2022, 29(2): 20-27.
[14]
KAPPEF, ZIRNGIBLC, SCHLEICHB, et al. Determining the Influence of Different Process Parameters on the Versatile Self-piercing Riveting Process Using Numerical Methods[J]. Journal of Manufacturing Processes, 2022, 84: 1438-1448.
[15]
FERRÁNDIZB, DAOUDM, KOHOUTN, et al. Prediction of Cross-sectional Features of SPR Joints Based on the Punch Force-displacement Curve Using Machine Learning[J]. The International Journal of Advanced Manufacturing Technology, 2023, 128(9): 4023-4034.
[16]
LIMushi, LIUZhao, HUANGLi, et al. Multi-fidelity Data-driven Optimization Design Framework for Self-piercing Riveting Process Parameters[J]. Journal of Manufacturing Processes, 2023, 99: 812-824.
[17]
ZHAOHuan, HANLi, LIUYunpeng, et al. Quality Prediction and Rivet/Die Selection for SPR Joints with Artificial Neural Network and Genetic Algorithm[J]. Journal of Manufacturing Processes, 2021, 66: 574-594.
TUXiaoming, JINJiageng, LIGuangyao, et al. Research on Matching Design of Electromagnetic Self Piercing Riveting Die Based on Neural Network Prediction[J]. Journal of Plasticity Engineering, 2025, 32(5): 43-51.
[20]
NEUKAMMF, FEUCHTM, BISCHOFFM. On the Application of Continuum Damage Models to Sheet Metal Forming Simulations[C]∥ International Conference on Computational Plasticity COMPLAS X. Barcelona, 2009:1-4.
[21]
LEES K, LEEJ S, SONGJ H, et al. Fracture Simulation of Cold Roll Forming Process for Aluminum 7075-T6 Automotive Bumper Beam Using GISSMO Damage Model[J]. Procedia Manufacturing, 2018, 15: 751-758.
[22]
TILLE T, HACKLB, SCHAUERH. Crash Simulation of Roll Formed Parts by Damage Modelling Taking into Account Preforming Effects[J]. AIP Conference Proceedings, 2011, 1383(1): 267-274.
[23]
BAOYingbin, WIERZBICKIT. A Comparative Study on Various Ductile Crack Formation Criteria[J]. Journal of Engineering Materials and Technology, 2004, 126(3): 314-324.
[24]
BAIYuanli, WIERZBICKIT. Application of Extended Mohr–Coulomb Criterion to Ductile Fracture[J]. International Journal of Fracture, 2010, 161(1): 1-20.
GANShouwu, CHENZhijun, ZHAOLeina. Construction of Flow Stress Prediction Model for Ti-55531 Alloy Based on BP Neural Network with Single Hidden Layer[J]. Hot Working Technology, 2018, 47(22): 59-61.
LINWeilu, DINGXiaofeng, SHUANGYuanhua. Prediction on Rolling Force of Oblique Rolling Piercing Based on BP Neural Network[J]. Forging & Stamping Technology, 2018, 43(10): 175-178.
[29]
程玥. BP神经网络的算法改进与应用研究[D]. 重庆: 重庆大学, 2011.
[30]
CHENGYue. Research on Algorithm Improvement and Application of BP Neural Network[D]. Chongqing: Chongqing University, 2011.
ZHANGBinghua, WANGXiaoliang. Rapid Content Determination of Vinpocetine Injections by NIR-BP United Quantitative Analysis Model[J]. Northwest Pharmaceutical Journal, 2022, 37(6): 30-36.
[33]
GAOXiaozhi, GUOWang, MEIChunxiao, et al. Short-term Wind Power Forecasting Based on SSA-VMD-LSTM[J]. Energy Reports, 2023, 9: 335-344.
[34]
XUEJiankai, SHENBo. A Novel Swarm Intelligence Optimization Approach: Sparrow Search Algorithm[J]. Systems Science & Control Engineering, 2020, 8(1): 22-34.
[35]
LIXingjia, GUJinan, SUNXiaohong, et al. Parameter Identification of Robot Manipulators with Unknown Payloads Using an Improved Chaotic Sparrow Search Algorithm[J]. Applied Intelligence, 2022, 52(9): 10341-10351.
ZHANGDamin, XUHang, WANGYirou, et al. Whale Optimization Algorithm for Embedded Circle Mapping and Onedimensional Oppositional Learning Based Small Hole Imaging[J]. Control and Decision, 2021, 36(5): 1173-1180.
[38]
YANGXinshe. Nature-inspired Metaheuristic Algorithms[M]. Second Edition. Frome: Luniver Press, 2010.
YANGJunqi, FANXiaojun, ZHAOYuehua, et al. Prediction of Carbon Emissions in Shanxi Province Based on PSO-BP Neural Network[J]. Journal of Environmental Engineering Technology, 2023, 13(6): 2016-2024.