摘要
傍山公路边坡区域发生滑坡、塌方等自然灾害的风险高,路侧设备的部署对行车信息服务与灾害监测至关重要。由于其远离电网,长距离供电经济性差,亟需场景匹配的供能方式。路域高熵能源的供能特性与设备用能需求相匹配,可在公路部署“高熵-荷-储”一体化设备,构成自供电微网,保障行车安全和公路应急响应能力。然而,高熵能源的捕获和利用在复杂公路场景下面临设备异常和出力随机性的挑战,这对自供电微网的正常运行带来一定影响。针对此问题,提出了一种基于高熵能源可用度评估的路域微网运行态势预警方法。首先,在设备层进行基于竞争失效的设备可靠性建模,并通过同类型多个设备间对比,提取出与安装条件无关的群体特征量,构建基于群体特征的深度学习模型,实时评估高熵能源捕获设备异常程度;其次,在集群层评估高熵能源集群输出是否达到预期捕获能量水平;然后,在微网层进一步评估高熵能源对微网负荷供电和储能功率支撑能力,提出基于支持向量数据描述的自供电微网运行态势预警方法;最后,通过算例验证基于可用度评估的微网态势预警方法有效性。研究结果表明:所提方法能够有效处理高熵能源运行状态评价过程中的不确定性,提高了评价结果的实时性与客观性,可有效识别自供电微网运行异常趋势,在线实现微网运行态势预警,保障微网的可靠运行。同时,成果可为高熵能源捕获设备运维管理、自供电微网运行态势预警提供了理论依据。
The sloped regions of highways situated in mountainous areas are particularly susceptible to natural disasters,including landslides and structural failures.Consequently,the installation of roadside equipment is essential for the provision of traffic information services and the monitoring of such disasters.Given the challenges posed by the remoteness from the power grid and the economic impracticality of long-distance power supply,there is an urgent need for an energy solution that is contextually appropriate.The energy attributes of high-entropy energy are well-suited to meet the energy requirements of the equipment utilized in these settings.The implementation of integrated high-entropy-load-storage systems along highways has the potential to create a self-powered microgrid,thereby enhancing traffic safety and improving emergency response capabilities on these thoroughfares.Nonetheless,the process of harnessing highentropy energy in intricate road environments presents several challenges,including equipment malfunctions and variability in energy output,which can negatively affect the functionality of the self-powered microgrid.To mitigate these challenges,this study introduces a methodology for monitoring the operational status of road-domain microgrid,predicated on the assessment of highentropy energy availability.Initially,at the equipment level,a reliability model that accounts for competitive failure is developed.Furthermore,the study extracts group characteristics that are independent of installation conditions by conducting comparative analyses of multiple similar devices.Utilizing the identified group characteristics,a deep learning model is developed to facilitate real-time evaluations of anomalies in high-entropy energy capture equipment.Subsequently,at the cluster level,the output generated by the high-entropy energy cluster is analyzed to ascertain whether it meets the anticipated energy capture thresholds.Furthermore,at the microgrid level,an assessment is conducted regarding the capacity of high-entropy energy to fulfill microgrid load demands and storage requirements.In this context,a method for situation warning of self-powered microgrid operational conditions,grounded in Support Vector Data Description,is proposed.Finally,the efficacy of the suggested microgrid situation warning approach,predicated on availability assessment,is corroborated through case studies.The findings of the study indicate that the proposed methodology is capable of effectively addressing the uncertainties associated with the evaluation of high-entropy energy operational status,thereby improving the timeliness and objectivity of the evaluation outcomes.It enables the effective identification of abnormal operational trends within self-powered microgrid and achieves online,proactive condition monitoring.This capability significantly enhances the operational reliability assurance of the microgrid.Furthermore,it establishes a theoretical foundation for the management of operation and maintenance of high-entropy energy capture equipment,as well as for the monitoring of operational conditions in self-powered microgrid.
作者
刘宝柱
廖钦伟
胡俊杰
朱兴一
LIU Bao-zhu;LIAO Qin-wei;HU Jun-jie;ZHU Xing-yi(School of Electrical and Electronic Engineering,North China Electric Power University,Beijing 102206,China;Key Laboratory of Road and Traffic Engineering of Ministry of Education,Tongji University,Shanghai 201804,China)
出处
《中国公路学报》
北大核心
2025年第11期224-240,共17页
China Journal of Highway and Transport
基金
国家重点研发计划项目(2023YFB2604600)。
关键词
交通工程
可用度评估
深度学习模型
路域高熵能源
态势预警
traffic engineering
availability assessment
deep learning model
road-domain highentropy energy
situation warning
