YAO Hong,YU Qilong,NA Lin*.Prediction Model for Corrosion Resistance of Phosphating Film Based on Particle Swarm Optimization Algorithm and Generalized Regression Neural Network[J].Plating & Finishing,2021,(11):1-6.[doi:10.3969/j.issn.1001-3849.2021.11.001]
基于粒子群优化算法-广义回归神经网络的磷化膜耐蚀性预测模型
- Title:
- Prediction Model for Corrosion Resistance of Phosphating Film Based on Particle Swarm Optimization Algorithm and Generalized Regression Neural Network
- 文献标志码:
- A
- 摘要:
- 选取磷化液温度、磷化液游离酸度和磷化时间作为输入参数，耐点蚀时间作为输出参数，引入广义回归神经网络（GRNN）建立磷化膜耐蚀性预测模型，并分别采用果蝇优化算法（FOA）、粒子群优化算法（PSO）对平滑因子寻优进而优化预测模型。使用18组训练样本对优化后模型进行训练，9组检验样本用于优化后模型的预测准确度评价。结果表明：PSO-GRNN模型的预测值非常接近真实值，预测相对误差在[0.001， 1.778]区间内，均方根误差最低、为0.682。与常规BPNN模型和FOA-GRNN模型相比，PSO-GRNN模型的预测准确度较高，对磷化膜耐蚀性预测效果良好。
- Abstract:
- The generalized regression neural network （GRNN） was introduced to establish the prediction model for corrosion resistance of phosphating film taking the temperature of phosphating solution, free acidity of phosphating solution and phosphating time as the input parameters and the pitting resistance time as the output parameter. Fruit fly optimization algorithm （FOA） and particle swarm optimization algorithm （PSO） were used to optimize the smoothing factor and then the prediction model was optimized. 18 groups of training samples were used to train the optimized model, and 9 groups of test samples were used to evaluate the prediction accuracy of the optimized model. The results showed that the predicted value of PSO-GRNN model was very close to the real value, the prediction relative error was within the range of [0.001, 1.778], and the root mean square error was the lowest of 0.682. Compared with conventional BPNN model and FOA-GRNN model, the prediction accuracy of PSO-GRNN model was higher, and the prediction effect of PSO-GRNN model for the corrosion resistance of phosphate film was excellent.
参考文献/References:
[1] 宗鹏洋, 王轶辰. 基于神经网络的软件质量评价综述[J]. 计算机科学, 2019, 46(11A): 507-516.
Zong P Y, Wang Y C. Software quality evaluation based on neural network: a systematic literature review[J]. Computer Science, 2019, 46(11A): 507-516 (in Chinese).
[2] 冯长敏, 张炳江. 基于BP神经网络的分段函数连续优化处理[J]. 北京信息科技大学学报(自然科学版), 2019, 34(1): 18-22.
Feng C M, Zhang B J. Continuous optimization of piecewise functions based on BP neural network[J]. Journal of Beijing Information Science & Technology University, 2019, 34(1): 18-22 (in Chinese).
[3] Wang H H, Wu Z L, Xing E P. Removing confounding factors associated weights in deep neural networks improves the prediction accuracy for healthcare applications[J]. Pac Symp Biocomput, 2019, 24: 54-65.
[4] Liu Z P, Chai X J, Chen X L. Deep memory and prediction neural network for video prediction[J]. Neurocomputing, 2019, 331: 235-241.
[5] 李娜, 刘冰, 王伟. 基于单隐层前馈神经网络的优化算法[J]. 科学技术与工程, 2019, 19(1): 136-141.
Li N, Liu B, Wang W. A optimization algorithm based on single hidden layer feedforward neural networks[J]. Science Technology and Engineering, 2019, 19(1): 136-141 (in Chinese).
[6] 王昕. 基于萤火虫算法-广义回归神经网络的光伏发电功率组合预测[J]. 电网技术, 2017, 41(2): 455-461.
Wang X. Combined PV power forecast based on firefly algorithm generalized regression neural network[J]. Power System Technology, 2017, 41(2): 455-461 (in Chinese).
[7] 陈超, 廖飞红, 周畅. 基于改进径向基神经网络的推力补偿算法[J]. 电子工业专用设备, 2020, 49(4): 1-5.
Chen C, Liao F H, Zhou C. Thrust force compensation algorithm based on improved radial basis function neural network[J]. Equipment for Electronic Products Manufacturing, 2020, 49(4): 1-5 (in Chinese).
[8] 张兆晨, 冀俊忠. 基于循环神经网络的时序fMRI数据分类方法研究[J]. 小型微型计算机系统, 2018, 39(7): 1426-1430.
Zhang Z C, Ji J Z. Research on the classification method of time-series fMRI data based on recurrent neural network[J]. Journal of Chinese Computer Systems, 2018, 39(7): 1426-1430 (in Chinese).
[9] 于晓明, 蒋静坪. 基于神经网络延时预测的自适应网络控制系统[J]. 浙江大学学报(工学版), 2012, 46(2): 194-198.
Yu X M, Jiang J P. Adaptive networked control system based on delay prediction using neural network[J]. Journal of Zhejiang University(Engineering Science), 2012, 46(2): 194-198 (in Chinese).
[10] 张翔, 李秋艺. SiO2颗粒分散液浓度对建筑结构钢锌系复合磷化膜耐蚀性的影响[J]. 电镀与精饰, 2020, 42(12): 10-14.
Zhang X, Li Q Y. Effect of concentration of SiO2 particles dispersion solution on corrosion resistance of zinc composite phosphating film on surface of building structural steel[J]. Plating & Finishing, 2020, 42(12): 10-14 (in Chinese).
[11] Hadzima B, Pastorek F, Borko K, et al. Effect of phosphating time on protection properties of hurealite coating: differences between ground and shot peened HSLA steel surface[J]. Surface and Coatings Technology, 2019(375): 608-620.
[12] Bensabra H, Azzouz N, Chopart J P. Effect of zinc phosphating treatment on the pitting corrosion resistance of steel reinforcement[J]. Revue de Metallurgie, 2013, 110(2): 153-163.
[13] 李龙澍, 张效见. 一种新的自适应惯性权重混沌PSO算法[J]. 计算机工程与应用, 2018, 54(9): 139-144.
Li L S, Zhang X J. New chaos particle swarm optimization based on adaptive inertia weight[J]. Computer Engineering and Applications, 2018, 54(9): 139-144 (in Chinese).
[14] 卢超, 杨翠丽, 乔俊飞. 基于PSO算法的动态模块化神经网络结构设计[J]. 控制与决策, 2018, 33(6): 1055-1061.
Lu C, Yang C L, Qiao J F. Dynamic modular neural network structure design based on PSO algorithm[J]. Control and Decision, 2018, 33(6): 1055-1061 (in Chinese).
备注/Memo
收稿日期： 2021-04-28；修回日期： 2021-06-30
作者简介： 姚宏（1981－），女，硕士，副教授，email：yao_20yao@126.com。
*通信作者： 那琳（1977－），男，硕士，副教授，主要研究方向：计算机应用技术。
基金项目： 河北省秦皇岛市科技局项目（201703A017）