本文已被:浏览 800次 下载 854次 |
码上扫一扫! |
|
基于无监督机器学习的胶州湾海底工程环境适宜性综合评价 |
杜星1,2, 孙永福3, 董杰4,5, 王青4,5, 宋玉鹏1,2, 苏志明1, 张莞君1
|
1.自然资源部第一海洋研究所 山东青岛 266061;2.青岛海洋科学与技术试点国家实验室 海洋地质过程与环境功能实验室 山东青岛 266235;3.国家深海基地管理中心 山东青岛 266237;4.青岛地质工程勘察院(青岛地质勘察开发局) 山东青岛 266100;5.自然资源部滨海城市地下空间地质安全重点实验室 山东青岛 266100
|
|
摘要: |
海岸带工程地质环境的稳定性对于海洋工程的建设安全和沿海经济繁荣十分重要。在胶州湾海域已有地质、水文等数据的基础上,对胶州湾海底工程环境适宜性进行了分区。通过无监督机器学习的谱聚类算法,构建了胶州湾海底工程环境适宜性综合评价模型。结果表明,胶州湾整体工程环境适宜性趋势为北高南低,从北向南依次可分为适宜性高、适宜性较高、适宜性较低和适宜性低四个区域。相关性分析表明,影响胶州湾海域海底工程适宜性的因素从高到低依次为冲淤分布、沉积物类型、坡度、第四系沉积物厚度、水深、海流流速、断裂分布。本研究可为胶州湾工程环境和地质灾害预防提供参考,有助于海洋工程环境稳定和经济安全保障。 |
关键词: 胶州湾 海底工程 环境适宜性 无监督机器学习 综合评价 谱聚类 |
DOI:10.11693/hyhz20210900206 |
分类号:P642.22 |
基金项目:海洋一所基本科研业务专项,GY0222Q05号;国家自然科学基金项目,42102326号;山东省自然科学基金项目,ZR2020QD073号。 |
|
ASSESSMENT AND SUBDIVISION OF ENVIRONMENTAL SUITABILITY FOR SUBMARINE ENGINEERING IN THE JIAOZHOU BAY BY UNSUPERVISED MACHINE LEARNING |
DU Xing1,2, SUN Yong-Fu3, DONG Jie4,5, WANG Qing4,5, SONG Yu-Peng1,2, SU Zhi-Ming1, ZHANG Wan-Jun1
|
1.The First Institute of Oceanography, MNR, Qingdao 266061, China;2.Marine Geology and Environment Laboratory Process, Pilot National Laboratory for Marine Science and Technology (Qingdao), Qingdao 266235, China;3.National Deep Sea Center, Qingdao 266237, China;4.Qingdao Geo-Engineering Exploration Institute (Qingdao Geological Exploration and Development Bureau), Qingdao 266100, China;5.Key Laboratory of Geological Safety of Coastal Urban Underground Space, Ministry of Natural Resources, Qingdao 266100, China
|
Abstract: |
The stability of the coastal engineering geological environment is essential for the construction safety of marine engineering and the prosperity of the coastal economy. Based on the existing geological and hydrological data in the Jiaozhou Bay waters, Qingdao, China, we divided the environmental suitability of the Jiaozhou Bay subsea engineering. Through the spectral clustering algorithm of unsupervised machine learning, a comprehensive evaluation model for the environmental suitability of Jiaozhou Bay subsea engineering was constructed. The results show that the overall engineering environmental suitability trend of Jiaozhou Bay is high in the north and low in the south. From north to south, it can be divided into four zones from high suitability to low suitability. Correlation analysis shows that the factors affecting the suitability of seabed engineering in Jiaozhou Bay are, from high to low, silting distribution, sediment type, slope, Quaternary sediment thickness, water depth, current velocity, and fault distribution. This study can provide reference for Jiaozhou Bay engineering environment and geological disaster prevention, and contribute to the environmental stability and economic security of marine engineering. |
Key words: Jiaozhou Bay submarine engineering geo-environmental suitability unsupervised machine learning comprehensive evaluation spectral clustering |