摘要
传统余热阀门控制技术主要分为机理建模和数据驱动2种方法,但在实际的应用中前者因机理复杂,难以准确描述,后者要求数据质量高、工况样本全,难以短时间满足。针对上述问题,提出一种基于融合驱动的余热阀门控制优化方法,该方法首先融合机理知识与数据知识构建基于模糊集合的知识图谱模型,将阀门开度知识实体化;其次,建立基于时间保护机制的长短时记忆(long short-term memory,LSTM)神经网络阀门开度优化模型,并提出时间保护机制算法,确定阀门最优调节频率;最后,通过知识推理得到推荐阀门开度。经实验分析验证,该方法通过融合余热回收机理等定性知识和设备运行数据等定量知识,在提升设备安全性的同时,产生的高温饱和蒸汽焓值提升概率为94%,平均每天可提升8640 kJ,实现了余热回收阀门开度的智慧决策。
The traditional waste heat valve control technology is mainly divided into two methods,mechanism modeling and data-driven.However,in practical applications,the former is difficult to accurately describe due to the complex mechanism.The latter requires high data quality and full working condition samples,which is difficult to meet in a short time.Aiming at the above problems,a fusion-driven optimization method for waste heat valve control is proposed.Firstly,the mechanism knowledge and data knowledge are fused to construct a knowledge graph model based on fuzzy sets,and the valve opening knowledge is materialized.Secondly,the LSTM valve opening optimization model based on time protection mechanism is established,and the time protection mechanism algorithm is proposed to determine the optimal adjustment frequency of the valve.Finally,the recommended valve opening is obtained by knowledge reasoning.Through experimental analysis and verification,this method integrates qualitative knowledge such as waste heat recovery mechanism and quantitative knowledge such as equipment operation data.While improving the safety of equipment,the probability of generating high-temperature saturated steam enthalpy is increased by 94%,and the average daily increase is 8640 kJ,which realizes the intelligent decision of waste heat recovery valve opening.
作者
刘晶
李超然
张建楠
赵佳
LIU Jing;LI Chaoran;ZHANG Jiannan;ZHAO Jia(School of Artificial Intelligence,Hebei University of Technology,Tianjin 300400,China;Hebei Data Driven Industrial Intelligent Engineering Research Center,Tianjin 300400,China;Tianjin Development Zone Jingnuo Data Technology Co.,Ltd.,Tianjin 300400,China;School of Science,Hebei University of Technology,Tianjin 300400,China)
出处
《热力发电》
CAS
CSCD
北大核心
2023年第10期176-186,共11页
Thermal Power Generation
基金
京津冀基础研究合作专项项目(G2021202013)
河北省自然科学基金资助项目(F2022202021)。