| 邓羽,张杰,彭中波,徐玮辰.基于人工神经网络的Ni-ZrO2纳米镀层耐腐蚀性能预测[J].装备环境工程,2022,19(2):98-105. DENG Yu,ZHANG Jie,PENG Zhong-bo,XU Wei-chen.Prediction of Corrosion Resistance of Ni-ZrO2 Nano-Plating Based on Artificial Neural Network[J].Equipment Environmental Engineering,2022,19(2):98-105. |
| 基于人工神经网络的Ni-ZrO2纳米镀层耐腐蚀性能预测 |
| Prediction of Corrosion Resistance of Ni-ZrO2 Nano-Plating Based on Artificial Neural Network |
| 投稿时间：2021-10-14 修订日期：2021-11-19 |
| DOI：10.7643/issn.1672-9242.2022.02.016 |
| 中文关键词: 电沉积 Ni-ZrO2纳米镀层 GRNN神经网络 BP神经网络 自腐蚀电流密度 预测 |
| 英文关键词:electrodeposition Ni-ZrO2 nano-plating GRNN neural network BP neural network self-corrosion current density prediction |
| 基金项目:国家自然科学基金(41376003)；中国科学院战略性先导科技专项(A类)(XDA13040405) |
|
|
| Author | Institution |
| DENG Yu | School of Shipping and Naval Architecture, Chongqing Jiaotong University, Chongqing 400074, China;Key Laboratory of Marine Environmental Corrosion and Bio-fouling, Institute of Oceanology, Chinese Academy of Sciences, Qingdao 266071, China;Open Studio for Marine Corrosion and Protection, Pilot National Laboratory for Marine Science and Technology Qingdao, Qingdao 266237, China |
| ZHANG Jie | School of Shipping and Naval Architecture, Chongqing Jiaotong University, Chongqing 400074, China;Key Laboratory of Marine Environmental Corrosion and Bio-fouling, Institute of Oceanology, Chinese Academy of Sciences, Qingdao 266071, China;Open Studio for Marine Corrosion and Protection, Pilot National Laboratory for Marine Science and Technology Qingdao, Qingdao 266237, China;Center for Ocean Mega-science, Chinese Academy of Sciences, Qingdao 266071, China |
| PENG Zhong-bo | School of Shipping and Naval Architecture, Chongqing Jiaotong University, Chongqing 400074, China |
| XU Wei-chen | Key Laboratory of Marine Environmental Corrosion and Bio-fouling, Institute of Oceanology, Chinese Academy of Sciences, Qingdao 266071, China;Open Studio for Marine Corrosion and Protection, Pilot National Laboratory for Marine Science and Technology Qingdao, Qingdao 266237, China |
|
| 摘要点击次数: |
| 全文下载次数: |
| 中文摘要: |
| 目的 对Ni-ZrO2纳米镀层的耐腐蚀性能进行预测，优化电镀工艺参数。方法 采用磁力搅拌辅助电沉积法，在钴镍基模型合金试样表面制备Ni-ZrO2纳米镀层，针对电镀工艺条件，设置正交实验，对每组实验镀层进行电化学测试，分析不同工艺条件下镀层的耐蚀性能。将ZrO2粒子浓度、电镀液温度和电镀电流密度作为神经网络的输入层，将自腐蚀电流密度作为输出层，运用GRNN神经网络和BP神经网络模型，对Ni-ZrO2纳米镀层进行耐腐蚀性能的预测研究。结果 当ZrO2粒子质量浓度为6 g/L、电镀液温度为60 ℃、电镀电流密度为5 A/dm2时，Ni-ZrO2纳米镀层的性能良好，表现出较小的自腐蚀电流密度。影响Ni-ZrO2镀层自腐蚀电流密度的因素满足ZrO2粒子浓度>电镀液温度>电镀电流密度。运用GRNN神经网络和BP神经网络对4组非正交实验预测的平均相对误差分别为5.30%与10.74%。结论 运用神经模型可以有效地预测不同工艺参数下镀层的耐腐蚀性能，从而优化工艺参数，提高实验效率。在训练样本较少的情况下，GRNN神经网络的预测性能更加精确。 |
| 英文摘要: |
| The work aims to predict the corrosion resistance of Ni-ZrO2 nano-plating and optimize the plating process parameters. Ni-ZrO2 nano-plating was prepared on the surface of cobalt-nickel-based model alloy samples by magnetic stirring-assisted electrodeposition. According to the electroplating process conditions, orthogonal experiments were set up, and each group of experimental plating was electrochemically tested to analyze the corrosion resistance with different process conditions. The three plating process parameters of ZrO2 particle concentration, plating solution temperature and plating current density were used as the input layer of the neural network, and the self-corrosion current density was used as the output layer. The GRNN neural network and BP neural network models were used to predict the corrosion resistance of Ni-ZrO2 nano-plating. When the mass concentration of ZrO2 particles is 6 g/L, the temperature of the plating solution is 60 ℃, and the plating current density is 5 A/dm2, the performance of the Ni-ZrO2 nano-plating is good, showing a small self-corrosion current density. The factors affecting the self-corrosion current density of Ni-ZrO2 nano-plating should meet the following requirements ZrO2 particle concentration>plating solution temperature> plating current density. The average relative errors of the four groups of non-orthogonal experiments predicted by GRNN neural network and BP neural network are 5.30% and 10.74%, respectively. The neural model can effectively predict the corrosion resistance of the plating under different process parameters, thereby optimizing the process parameters and improving the experimental efficiency. In the case of fewer training samples, the prediction performance of the GRNN neural network is more accurate. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
