| 张彭辉,彭文山,刘少通,丁康康,侯健.基于深度森林算法的船用钢腐蚀速率预测模型构建[J].装备环境工程,2025,22(4):117-124. ZHANG Penghui,PENG Wenshan,LIU Shaotong,DING Kangkang,HOU Jian.Establishment of Hull Steel Corrosion Rate Prediction Model Based on Deep Forest Algorithm[J].Equipment Environmental Engineering,2025,22(4):117-124. |
| 基于深度森林算法的船用钢腐蚀速率预测模型构建 |
| Establishment of Hull Steel Corrosion Rate Prediction Model Based on Deep Forest Algorithm |
| 投稿时间：2025-02-18 修订日期：2025-03-10 |
| DOI：10.7643/issn.1672-9242.2025.04.015 |
| 中文关键词: 深度森林 预测模型 船用钢 海水腐蚀中图分类号:TG172 文献标志码:A 文章编号:1672-9242(2025)04-0117-08 |
| 英文关键词:deep forest prediction model hull steel seawater corrosion |
| 基金项目: |
|
|
| Author | Institution |
| ZHANG Penghui | National Key Laboratory of Marine Corrosion and Protection, Luoyang Ship Material Research Institute, Shandong Qingdao 266237, China |
| PENG Wenshan | National Key Laboratory of Marine Corrosion and Protection, Luoyang Ship Material Research Institute, Shandong Qingdao 266237, China |
| LIU Shaotong | National Key Laboratory of Marine Corrosion and Protection, Luoyang Ship Material Research Institute, Shandong Qingdao 266237, China |
| DING Kangkang | National Key Laboratory of Marine Corrosion and Protection, Luoyang Ship Material Research Institute, Shandong Qingdao 266237, China |
| HOU Jian | National Key Laboratory of Marine Corrosion and Protection, Luoyang Ship Material Research Institute, Shandong Qingdao 266237, China |
|
| 摘要点击次数: |
| 全文下载次数: |
| 中文摘要: |
| 目的 对深度森林算法在材料海水环境腐蚀数据预测方面的应用进行研究。方法 对深度森林算法的原理进行介绍，以船用钢在我国近海海域的腐蚀试验数据为样本进行应用，构建预测模型，并对模型性能进行评价。结果 应用深度森林算法构建的模型与传统神经网络算法模型相比，其性能和预测准确度均较高，并具有很好的泛化能力。结论 深度森林算法模型具有较好的预测准确性和通用性，可满足海水环境腐蚀数据预测应用需求。 |
| 英文摘要: |
| The work aims to study the application of the deep forest algorithm in predicting corrosion data of materials in seawater environments. In this paper, the principle of the deep forest algorithm was introduced, and applied to the establishment of a corrosion rate prediction model using the test data of hull steel obtained in China's offshore seawater to assess the performance of the model. Compared with the model constructed by applying the traditional neural network algorithm, the model constructed by applying the deep forest algorithm had higher performance and prediction accuracy, and had a very good generalization ability. The results showed that the model has good prediction accuracy and versatility, and can meet the application requirements for prediction of corrosion data of materials in seawater environments. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
