基于数据挖掘的耙吸船能效优化智能决策管理系统  

Intelligent decision management system of dredger energy efficiency optimization based on data mining

作　　者：陈力恒 张金越 王辉[1] 赵玉新 CHEN Liheng;ZHANG Jinyue;WANG Hui;ZHAO Yuxin(College of Intelligent Systems Science and Engineering,Harbin Engineering University,Harbin 150001,China;Jilin Jiangji Special Industry Co.,Ltd.,Jilin 132021,China)

机构地区：[1]哈尔滨工程大学智能科学与工程学院,黑龙江哈尔滨150001 [2]吉林江机特种工业有限公司,吉林吉林132021

出　　处：《实验技术与管理》2025年第2期44-51,共8页Experimental Technology and Management

基　　金：国家自然科学基金(62273114);黑龙江省自然科学基金(YQ2024F015);教育部产学合作协同育人项目(23110767293840);哈尔滨工程大学教学改革研究项目(JG2023B0407)。

摘　　要：耙吸船是一种大型自航、装仓式挖泥船,被广泛应用于港口与通航河道的疏浚工程。然而,耙吸船挖泥作业过程具有典型的非线性、迟滞性等特点,难以建立船舶能效与作业参数之间的动态响应关系,这使得目前缺乏对能效优化问题的理论分析与定量评估。针对上述问题,该文首先对“新海虎8号”耙吸船历史数据的特征规律进行挖掘与分析,通过深度学习方法建立能效预测模型,进而利用粒子群优化策略实现船舶能效的动态智能决策,在此基础上,该文开发了一套集耙吸船多源信息监测、存储与优化辅助决策为一体的智能能效管理系统。通过“新海虎8号”实船数据的分析与对比,验证了该文所提方法可以将耙吸船能耗节省20%以上,达到绿色航运与节能减排的目标。同时,该文所开发的能效优化智能决策管理系统可以作为人工智能专业核心课程“大数据信息挖掘”的典型工程案例与演示平台,使学生深入理解数据挖掘建模过程与方法,为相关研究奠定工程实践基础。[Objective]Trailing suction hopper dredgers(TSHDs)are a vital type of construction vessel in the modern dredging industry.Renowned for their maneuverability,efficiency,and ability to withstand strong winds and waves,TSHDs play an essential role in tasks like port maintenance and coastal channel excavation.However,their large operational volumes,high power demands,and significant energy consumption complicate traditional operation and optimization management.Additionally,the dredging operations of TSHDs are highly nonlinear and exhibit delayed responses,making it difficult to establish a dynamic response relationship between ship energy efficiency and operational parameters.This has limited theoretical analysis and quantitative evaluations aimed at optimizing energy efficiency.Currently,the design and implementation of energy efficiency management systems for dredging vessels remains incomplete.[Methods]This paper first applies big data analysis techniques to the historical data of the TSHD“New Sea-tiger 8,”examining the relationships between ship speed,dredger head parameters,and energy efficiency.This analysis lays a foundation for intelligent decision-making in energy efficiency optimization.Subsequently,deep learning techniques are employed to establish a prediction model for TSHD energy efficiency,considering the interplay of multiple influencing factors.Particle swarm optimization strategies are then utilized to optimize the energy efficiency of dredging operations.Finally,the big data analysis and intelligent optimization models and algorithms are integrated into an intelligent energy efficiency management system.This system collects and displays multisource data,utilizes self-learning,and performs rolling optimizations to enable real-time intelligent management of the dredger’s operations.[Results]The correlation analysis revealed that parameters like mud pump speed and mud pump motor speed had a strong correlation with energy efficiency,with absolute correlation coefficient values exceeding 0.8.This

关 键 词：疏浚耙吸船 长短期记忆神经网络 能效预测 粒子群优化 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论]